阅读说明：本技术 基于可穿戴压力传感器阵列的人体动作和位置感测、识别与分析 (Human body action and position sensing, identification and analysis based on wearable pressure sensor array ) 是由 本杰明·亚瑟·巴佐尔 朱子杰 潘挺睿 于 2018-10-09 设计创作，主要内容包括：基于压力传感器的阵列集成到控制装置中,控制装置检测用户一个或多个身体部位的位置、动作或运动,以识别动作并将其转换为独特的用户动作分布。用户动作分布可以被独立地分析或识别为离散的运动或姿势,并且可以用作控制装置本身的输入或命令,或者用作将输出信号传递给配套装置的信号或信号集。压力传感器可以连接到用户的任何身体部位,例如用户的手腕或脚踝。用户的动作或位置或其变化生成可用于控制配套装置的输出信号。可检测信号的来源是基于压力的传感器阵列,传感器阵列产生压力数据分布,压力数据分布转换为输出信号以控制配套装置。(The array of pressure-based sensors is integrated into a control device that detects the position, motion, or movement of one or more body parts of the user to recognize the motion and convert it into a unique user motion profile. The user motion profile may be independently analyzed or recognized as a discrete motion or gesture, and may be used as an input or command to the control device itself, or as a signal or set of signals to communicate an output signal to a companion device. The pressure sensor may be attached to any body part of the user, such as the wrist or ankle of the user. The user's motion or position or changes thereof generate output signals that can be used to control the fitting. The source of the detectable signal is a pressure-based sensor array that produces a pressure data profile that is converted to an output signal to control the companion device.)

1. An apparatus for detecting user actions, comprising:

a pressure sensor array comprising a plurality of individual pressure sensors spaced around a support band having a tensioning element for maintaining the sensors in contact with the skin around the user's limb, wherein the plurality of sensors are spaced around the band to detect pressure changes caused by the user's actions; and

a pressure data profile resulting from the user action detected by the sensor array, wherein the pressure data profile comprises a plurality of skin pressure measurements.

2. The apparatus of claim 1, wherein the array of pressure sensors comprises at least six individual pressure sensors arranged on an inner surface of a band sized to circumferentially surround a wrist of a person.

3. The apparatus of claim 1, further comprising a data storage unit containing the pressure data profile from the user, the pressure data profile comprising changes in pressure data profile associated with a particular gesture of the user.

4. The apparatus of claim 1, wherein at least four individual sensors are spaced around the band in a direction corresponding to the location of at least four separate tendons in a wrist of the person.

5. The apparatus of claim 3, wherein the pressure data profile comprises pressure sensor array data associated with motion of the user's arm and finger.

6. The apparatus of claim 1, wherein the pressure data profile comprises pressure measurements with a sensitivity of less than 1 mmHg.

7. The apparatus of claim 1, further comprising a separate sensor selected from the group consisting of an optical sensor, a gyroscope, a magnetometer, and an accelerometer, and combinations thereof, wherein the pressure data profile further comprises data from the separate sensor.

8. The apparatus of claim 2, wherein each sensor of the array is an ion sensor having a height of less than 1.5 mm.

9. The device of claim 1, further comprising a companion device selected from a digital watch, a digital telephone, a computer, a video game controller, a digital locking mechanism, or a combination thereof.

10. A method of translating a user's recognized gesture into a control signal for a companion device, comprising:

generating a pressure data distribution input resulting from the recognized gesture of the user detected by a pressure sensor array comprising a plurality of individual pressure sensors in skin contact around a limb of the user,

correlating the pressure data distribution input from the recognized gesture to an output signal;

sending the output signal to a companion device to cause the companion device to perform a function associated with the recognized gesture of the user.

11. The method of claim 10, wherein generating a pressure data profile comprises:

at least six individual pressure sensor values are measured, the pressure sensor values being for pressure sensors arranged circumferentially around and in contact with the skin of a person's wrist.

12. The method of claim 10, wherein the associating step comprises:

comparing the pressure data profile input to a stored pressure data profile corresponding to the recognized gesture of the user.

13. The method of claim 11, wherein comparing the input pressure data profile to the stored pressure data profile comprises:

comparing the pressure change caused by the first motion of the user's finger and tendon with the pressure change caused by the second motion of the user's finger and tendon.

14. The method of claim 11, wherein the input pressure data profile comprises pressure measurements from at least four sensors at skin adjacent at least four individual tendons of the user, the at least four sensors spaced around a band that remains fitted to the wrist of the user.

15. The method of claim 11, wherein the input pressure data profile has a sensitivity of less than 1 mmHg.

16. The method of claim 11, wherein the input pressure data profile further comprises an input of a separate sensor selected from the group consisting of an optical sensor, a gyroscope, a magnetometer, and an accelerometer, and combinations thereof.

17. The method of claim 11, wherein the generating an output signal for controlling the companion device comprises:

generating a data component that distinguishes a single user from other users such that the companion device will not function without the data component of the single user.

Background

A wide variety of electrical and mechanical devices may be coupled to a separate control device that receives input from a user that is sensed, recognized, and analyzed to interpret the input from the user and convert the input into an output that can be used for various purposes. Inputs from the user's body sensed and recognized by the control device may be communicated to the companion device, and such sensing, recognition and analysis processes enable the user to manually manipulate the control device to create a data output. This data output can be used for independent analysis of user actions or converted into commands or control instructions for the companion device. Examples include keyboards, touch pads, computer mice, microphones, numeric keypads, pedals, and various other commonly used input devices that are typically operated using hands or feet. In these examples, an action imparted to the device by the user (e.g., by tapping on various keys of a keyboard, moving a computer mouse, inputting a sound into a microphone, or activating a pedal) results in an output that instructs the companion device, e.g., as a computer, to perform certain steps. As is apparent from these examples, coupling a user's specific inputs to a desired output to a companion device is an integrated process that aims to translate the most convenient way a user provides input into data output necessary for the companion device to function properly.

The wide variety of input devices reflect a wide variety of companion devices that may be controlled through manual input from a user, including but not limited to computers, telephones, video displays, control and safety systems, and virtually any device or system having a control mechanism or interface through which a user directs control of the device.

Most users are familiar with input devices such as keyboards and numeric keypads in which manual contact with mechanical keys or buttons or a touch screen or space in the field of view is translated into a single letter, number or other instruction, such as turning the device on or off or causing a companion device to perform some predetermined function. In order to make these input devices faster, more efficient and more convenient, several motion or movement detector apparatuses have been developed which can be physically attached to a body appendage such as a user's hand, wrist or foot, and which can translate the user's motion or posture into data input or instructions, even if the motion or posture is performed entirely in "space", i.e. where the user is not in direct contact with any input device.

U.S. patent publications 2016/0091980 (apple Inc.) and 2017/0031453 (Philips Inc.), both of which are specifically incorporated herein by reference, describe devices that use light sensors to detect hand motion that rely on a pair of light emitters and a series of light detectors to identify hand motion based on the difference in light passing through the wrist anatomy when the hand is making certain gestures. These devices are intended to detect gestures based on the detection of light and then convert the light signals into data input. For example, by placing a light emitting device and a light detection sensing device around the wrist, which devices sense differences in light transmitted through bones, muscles and tendons, the user moves his hand, and based on the motion of the user's hand and wrist, the device converts the differences in transmitted light into instructions that can control a watch, computer or other device. For example, the system may detect hand waving motions of the user to indicate certain actions, such as turning the computer on or off, and may detect differences in light transmitted through tissue during various finger motions of the user to convert different finger motions into discrete signals to control the companion device.

These existing devices are typically coupled to a data processing unit that converts the sensed light signals from the user's motion or posture into specific output signals. For example, the combination of the sensor and the data processing unit may detect that when the user extends one finger, the optical properties of the user's bones, muscles and tendons differ with respect to extending two fingers, and recognize them as different signals, and then instruct the companion device to perform different functions. Thus, the hand and finger motion variations between extending one finger and extending two fingers may be different, for example, "turning" on "or" off "a computer or cell phone. These devices may also be combined with motion sensors and electrical sensors to generate mixed signal inputs or with light emitters and light sensors to reflect multiplexing (multiplex) of user actions, motions or gestures.

Although devices based primarily on light sensing arrays can distinguish between multiple individual movements of a user, these devices have certain inherent disadvantages in using light or electrical signals to detect movement, including the inherent possibility that background noise in the form of extraneous light or electrical signals compromises the accuracy of the detected signals. Furthermore, light emitting devices tend to require significant power to operate, and this power requirement can result in the need for associated large and bulky power storage devices, or in a limited lifetime of any device based on light or electrical sensors.

Disclosure of Invention

The present invention is a pressure sensor-based array integrated into a control device that detects the position, motion, or movement of one or more body parts of a user to identify the motion and convert it into a unique user motion profile. The user motion profile may be independently analyzed or recognized as a discrete motion or gesture, and may be used as an input or command to control the device itself, or as a signal or set of signals to generate an output signal to a companion device. The pressure sensor may be attached to any body part of the user, but is preferably attached to the user's wrist, ankle or other body part that may be surrounded by a plurality of sensors forming an array and is typically used to provide input to a companion device. For example, pressure sensor-based devices attached to the wrist detect a number of individual components based on basic physiological structures including movement of muscles, tendons and bones in the user's wrist, hand, arm and finger to convert position, motion or movement into detectable signals that can be used to generate output signals that are decoded to control a companion device. The source of the detectable signal is an array of pressure-based sensors attached to the user's limb, for example, an array attached to the wrist that detects individual movements of muscles, tendons and bones in the same manner as an array attached to the ankle that detects movement of muscles, tendons and bones of the foot or toe to generate a signal that produces an output signal that controls the companion device.

For ease of reference, the term "motion" is used hereinafter to describe each of sensing an initial, stationary or baseline position of a portion of a user's body as a first position, sensing a transitional motion (e.g., representing a different motion or gesture commanded) away from the first position, and sensing a final or second position different from the first position, wherein the difference between the first position and the second position and/or the transitional motion is interpreted as a signal or input, preferably for controlling a companion device. Thus, the device of the present invention can sense motion as a course of activity or detect differences in the body between initial and final positions or a combination of both as the user moves, recognize these motions or changes in position, and generate specific motion profiles for analysis, including conversion into motion profiles, for analysis when interpreted repeatedly as controls for a companion device.

Functionally, the device of the present invention detects pressure at multiple points along and/or around a user's body part and makes unique pressure measurements produced by the motion to convert pressure readings into a unique quantitative pressure profile that is analyzed and converted into any one or several discrete data profiles, including but not limited to motion profiles, discrete gestures, and instructions to a companion device. For example, pressure sensors are positioned circumferentially and detect pressure values along certain selected points of the inner surface of the band attached to the wrist that are characteristic of the movement of the forearm, hand, wrist, finger, any single cell or combination thereof, including the movement of a single finger, including numbers or letters in single and collectively designated alphanumeric tables. Different movements of the body, which can be recognized and correlated with gestures converted into outputs, such as commands for controlling a computer, by using a plurality of pressure measurements individually and collectively, generate different characteristic pressure values and combined distributions. Due to the sensitivity and selectivity of the particular pressure sensor used in the present invention, the characteristic motion of the body part can be used to distinguish subtle motions performed by the user, identify, output as a unique quantitative data distribution, analyze as a particular action or range of actions, and optionally assign specific command and control functions to the companion device.

To maximize the information obtained from the user's actions, a plurality of individual pressure sensors are combined to create an array of pressure sensors that can be combined with different types of sensors designed to detect other parameters in the same or different areas of body tissue, including selected portions of tendons, ligaments, muscles, bones, interstitial tissue, veins, arteries, and any body part that can give rise to static or differential measurements in addition to pressure readings of the point where the individual pressure sensors are in contact with the skin surface. The combination of the sensor array and additional sensors may also include a detector of pulse blood flow or other physiological parameters, as well as sensors of acceleration, rotation or changes in electric or magnetic fields. In general, the signals from all sensors create a data output and a sensor profile, wherein each movement of the user's limb generates a specific, quantitative pressure profile that is unique to a single action and optionally to an individual. Thus, the motion of a user extending a single index finger is easily distinguished from the motion of a user extending an index and middle finger, and the unique specific pressure sensor profile for each motion or gesture produces a discrete command or data output that is preferably processed into a pressure-based data profile that is analyzed and can be converted into inputs or commands and control instructions for a companion device.

When such a sensor array comprising an array of pressure sensors is deployed as described herein, a local, optionally circumferentially oriented "map" of pressure data may be constructed for any motion of any limb or body part. As the number and density of pressure sensors increases, as well as the number of combinations of individual pressure data values increases, more information about the location or movement of the underlying body structure will be collected for subsequent analysis and generation of output instructions to the companion device. In the case of a single sensor across the circumference of the wrist, the strain (related to the expansion/contraction of the wrist, such as the flexion and extension of the hand) can be determined from the single and collective motion of the underlying physiological structures.

As the number of sensors increases and the size of each sensor decreases, the size of the sensors can approach the average tendon size (about 4 mm). Thus, as the number of sensors increases (e.g., over an array of 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 32, or 64 individual sensors), the number of sensors will exceed the number of underlying physiological structures to which pressure is detected, and the sensor data combination is made up of discrete measurements of multiple individual anatomical structures.

Thus, two adjacent sensors may each detect a pressure contribution from any one or more underlying anatomical structures, including but not limited to one or more tendons, one or more muscles, one or more ligaments, and/or one or more bones. When the number of sensors approaches twice the number of measured physical features (e.g., combinations of tendons, muscles, bones, etc., for a total of about 18 features), individual sensors more directly measure individual physiological structures, and specific movements of the limb can be tracked more accurately using individual physiological structures or combinations of individual physiological structures in combination with multiplexing (multiplexing) of sensor data from other individual physiological structures or multiple individual physiological structures. A signal processing technique called de-aliasing allows the separation/determination of changes affecting multiple sensors. Similar techniques can accommodate variations in sensor rotation about the wrist. Where the array includes fewer sensors, the sensed physiological structures, the total number of tendon groups (e.g., 8 sensors in the array arranged around more than 8 tendons), can be tracked, most natural hand positions (including but not limited to hand flexion/extension, adduction/abduction, and various finger flexion/extension patterns) can be determined

The control device may have a dedicated power supply and circuitry for connecting each pressure sensor to at least one data storage medium, and optionally logic for identifying and analyzing specific pressure profiles, and the signal processing for the quantitative unique pressure profile may be contained in the control device or a companion device. Thus, a control device attached to the wrist may process the command and control functions resulting from the pressure profile, or the pressure profile may be transmitted to a device such as a computer, telephone, game controller or other companion device containing logic to convert the motion into command and control functions. Preferably, data is collected at least three, preferably five or more critical pressures. For example, a particular sensor may detect a zero change in direction or motion, a small positive change in pressure resulting from motion, a large positive change in pressure resulting from motion, a small negative change in pressure, or a large negative change in pressure. In addition to discrete quantitative values, individual thresholds may be preset to combine unique quantitative pressure profiles as described herein.

By combining data from multiple individual sensors deployed in an array, individual pressure profiles corresponding to particular individual or discrete actions, ranges of actions, sets of combined actions, or subsets of actions can be identified to establish patterns of actions or deviations from existing patterns. Any absolute or relative measure of motion, range, or pattern of motion may be analyzed and used to generate a quantitative score that may reflect individual or combined/multiple pressure measurements, individually or collectively, and compared to a previous individual motion or set of motion combinations. The individual or collective action scores may be associated with types of actions performed by the user that are characteristic of functions to be performed by the user, or characteristics of controls or commands to be controlled by the user to a companion device.

If the user associates a pressure profile with a particular gesture, each gesture may be assigned a command or control function for controlling the device or companion device. For example, a set of pressure changes detected on at least six tendons may indicate a particular gesture corresponding to a particular command output. Similarly, slight changes in four or more tendons plus zero changes in two tendons may indicate different specific gestures and different specific command outputs. In practice, the set of muscles, tendons, ligaments, and bones' directions and actions established when the user makes a fist may instruct the companion device to "power on," while the set of unique muscles, tendons, and bones established when the user indicates the index finger may instruct the companion device to turn on a particular program or turn off power. As described in detail below, a large number of movements and gestures may be measured and assigned different commands and control instructions by pre-programming or learning modes at the direction of the user. All of these functions, as well as further functions, will be understood by those of ordinary skill in the art in light of the following description and the accompanying drawings.

Brief description of the drawings

Fig. 1 is a graph showing six independent metrics (G0-G5) of each element of a quantitative posture detection data distribution of the control device of the present invention. Metric combinations based on the individual magnitudes of the individual metrics produce unique quantitative gesture detection scores based on the particular individual pressure values input from the sensor array and reflected in metrics G0-G5.

FIG. 2 is a cross-sectional view of a human wrist showing the orientation of various physiological structures including the 14 tendons, radius and ulna, and other internal anatomical structures, such that placement of multiple individual pressure sensors conforms closely to the outer skin layer of the wrist, resulting in a unique quantitative pressure distribution based on the different configurations and actions of the forearm, wrist, hand and fingers.

Fig. 3A and 3B are control devices of the present invention conforming to a user's limb showing the orientation of tendons in the physiological structure of a human wrist with sensors placed in close proximity to the tendons.

FIG. 4 is a combined abnormal human gesture consisting of arm movement, finger movement, and palm movement.

FIG. 5 is a combination of gestures and commands that may be correlated with pressure sensor data, and illustrates an example of how certain gestures may be translated into commands for three different controllers.

Fig. 6 is a graph showing representative changes in physiological interface pressure for fourteen individual tendons located in the right wrist. The blank field indicates no change in pressure value, the single downward arrow indicates a small decrease in pressure value, the double downward arrow indicates a large decrease in pressure value, the single upward arrow indicates a small increase in pressure value, and the double upward arrow indicates a large increase in pressure value. The left axis contains at least 11 individual directions or command and control functions that can be assigned to a particular combination of measured tendon pressure changes.

Detailed description of the invention

The present invention relates to a pressure sensor-based control device including individual pressure sensors forming an array, a control device in combination with a companion device, and methods of using the same, which use a plurality of pressure sensors to detect user actions to generate data, data distributions, quantitative pressure signals, defined gestures, control instructions, commands, or other inputs directed to the control device or companion device based on sensor inputs of the plurality of sensors forming the array. The control device is physically attached to the user's limb by any structural or mechanical means that firmly positions the array to the limb, such that the pressure-based sensor is capable of detecting pressure generated by subcutaneous anatomical structures in either of a first configuration or an initial resting state, a series of movements comprising a transitional motion state or range of states, and after the transitional motion, generating a second state that is different from the first configuration, and may follow any number of different configurations.

Although the sensor input relies on multiple pressure sensors, additional optical sensors, acceleration or inertial sensors, gyroscopic or other rotational sensors, and magnetic sensors may independently provide input related to the user's motion, and may be combined with pressure sensor data as described herein. In some embodiments, data from the array of pressure sensors is integrated with data from at least 3 of each of the accelerometer, gyroscope, and magnetometer.

Based on the detected motion, the control device generates an input to the control device itself or the companion device, and the control device and/or the companion device performs an operation based on the input. The input from the control means may be combined with other conventional user interface mechanisms such as a keyboard, keypad, touch screen, etc. so that the control means of the present invention works in conjunction with existing input devices. Examples of companion devices include, but are not limited to, computers, cell phones, video display devices, games, motion or other interactive consoles, robotic motion and other remote control systems, musical instruments, medical devices, automobiles, appliances, and virtually any electronic or mechanical device capable of receiving input to control the state or operation of a companion device.

In some examples, the plurality of pressure sensors forming the pressure sensing array are located on the control device itself or on an accessory to a companion device containing the sensor array. For example, the sensor array may be located in an integrated assembly of the control device and the watch as a companion device, or may be a discrete device separate from the watch, but may be operably connected by any conventional communication mechanism, such as bluetooth, a wired connection, optical or wireless transmission, or any other commonly used data transmission mechanism or method.

The array of pressure sensors produces a unique specific pressure profile based on the subcutaneous pressure applied by the user's tendons, ligaments, skin, muscles and bones. In some embodiments, the control device is comprised of a functional sensing material that is typically used in clothing or other items of his wear that are attached to the user, such as watches, hats, jewelry, bandages, or other structures that bring the array of pressure sensors into close conforming engagement with the skin surface, maintaining a substantially uniform orientation on the user's body surface such that absolute or relative motion of the user causes a change in the particular pressure profile sensed by the control device.

Example 1-pressure sensor based control device is integrated into a conventional belt material (fabric, leather, silicone, polymer, metal or combination thereof).

In this example, the array of pressure sensors is embedded in the wristband, and the wristband serves as the control device. The sensors circumferentially surround the wrist and, because of the absence of the rigid sensor assembly, the wristband remains in conforming fit with the user's skin surface, as does a typical wristband. The individual pressure sensors, which are flexible, form an array and are integrated or embedded as part of the watchband material such that the watchband area includes an array of sensors that cover a substantial circumference of the wrist or are not occupied by a control device or a companion device. Thus, where the sensing array covers the entire wristband, the array spans the entire wrist, such that the sensors are positioned around the entire circumference of the limb, this configuration maximizes the sensing input from each pressure applied to the skin surface, and maximizes the data as the number of sensors increases. As will be readily appreciated, a greater number of sensors may be arranged around a smaller arc around the limb and, depending on underlying physiology, the resulting data distribution will depend on the area covered by the sensors and the number of sensors arranged in the array. Preferred embodiments of the number of sensors and the coverage area around the body limb are described below and in the drawings.

Viewing the watchband along an axis transverse to the center of the wrist, the angle defines a discontinuity in the perimeter in which each sensor is deployed. For example, if the sensor is deployed around the entire perimeter of the watchband, the angle will be 360 °. Similarly, if a portion of the wristband is occupied by the case of the control device or the companion device such that each sensor is disposed about three-quarters of the circumference of the wristband, the angle would be 270 °. The particular angles contemplated by the pressure sensor array of the present invention include angles greater than 90 °, greater than 120 °, greater than 150 °, greater than 180 °, greater than 210 °, greater than 240 °, greater than 270 °, greater than 330 °, and integrated values thereof.

Example 2-sensor specification.

The single sensor preferably comprises an ion sensor as described in U.S. patents 9,170,166, 9,459,171, 9,739,679 (and co-pending U.S. application 2017/0059434a1 and the Fabric sensor application filed on 25/5). The sensor is preferably textile based, thin (500 μm or typically 1.5mm) and conforms to the outer surface of a curved surface of the body, such as the wrist, forearm, ankle, skull, neck, chest or abdomen. The sensor may also be integrated into a garment and customized for size, material and sensitivity depending on the application.

Each sensor must have a working pressure range to detect the maximum pressure change produced by the gesture and take into account the baseline pressure caused by the strap tension (range 0-100 mmHg). The biasing structure may change the minimum operating pressure from 0 and maintain the size of the maximum sensitivity range (e.g., 40-70mmHg for a region of high baseline pressure). A working range of 0-30mmHg (with offset configuration) is desirable, but can be as low as 0-10mmHg or as high as 0-120 mmHg. For basic gesture detection, pressure changes as low as 4mmHg must be detected and distinguished from noise, so the sensitivity (pressure resolution) and noise level is no greater than 1mmHg, and repeatability errors are no greater than 50%. For advanced posture detection, pressure changes as low as 0.5-1mmHg must be detectable, so the sensitivity and noise level is no greater than 0.2mmHg, most preferably in the case of 1Pa (0.0075mmHg), most preferably the repeatability error is below 10%. In the case of position sensing (rather than transition/motion sensing), accurate pressure readings are required. Considering accuracy errors from noise, linearity and repeatability, the system must accurately quantify tendon pressure to 3 to 5 classes. The accuracy is preferably ± 5mmHg (83.3% of the full scale range) for detecting the base hand position, and ± 1mmHg (96.7%) for the advanced hand position, and it is most preferable that the accuracy is higher than ± 0.5mmHg (98.3%). Linearity and repeatability accuracy must exceed the overall accuracy requirement. Thus, the accuracy of linearity and repeatability is preferably 90% for the base position, 98% for the premium position, and most preferably 99.5%.

The signal-to-noise ratio is greater than 100: 1 (equivalent to 0.3mmHg) is preferred, whereas 1000: a ratio of 1 is most preferred.

The sensor array preferably has a total vertical height perpendicular to the surface of the user's skin of 0mm (conformal contact), preferably no more than 0.5mm, most preferably no more than 1.1 mm. The control device has an FPC type connector, preferably with a plurality of positions related to the number of sensors (Sn). In some embodiments, the locations are Sn +1 (e.g., 9 locations of 8 sensors), while the other locations are the nearest integer greater than or equal to twice the square root of Sn (ceiling (2 × sqrt (Sn)) (e.g., 8 locations of 16 sensors). The ideal pin spacing is 1mm, ranging from 0.25mm to 2.54 mm.

Pressure data using the aforementioned pressure sensor can be acquired at low power. Power consumption is proportional to the sampling rate. 125Hz and 16mA are typical representatives of high performance pressure sensing systems. The theoretical maximum for these sensors is 240Hz due to the response time. In the case of postural monitoring, this is excessive. Human motion is typically less than 1Hz (1 second order change) and rarely exceeds 10Hz (100ms order change). Humans consider changes near/below 100ms to be near instantaneous and changes above 50Hz (20ms) to be instantaneous. A sampling frequency of 10Hz may produce a current of 1.6 mA. Together with inertial sensors (gyroscopes, accelerometers and magnetometers), the total current is typically 2.6 mA. The battery capacity of the wearable watch is 100-200mAh, so the running time is 42-78 hours or 8 hours of continuous operation of 5-9 days. Power saving functions (e.g., sleep mode) can be greatly expanded.

Due to the high signal-to-noise ratio (SNR) of the data, almost no processing is required for gesture acquisition (typically simple arithmetic operations) and the power consumed is negligible. EMG signals require wavelet analysis, strong analog amplification, and CPU intensive noise reduction, which consumes more power. For reference, a Myo EMG cuff charge may run for 1 day, consuming 2-4mA per channel (16-32 mA for 8 channels) by the modular EMG unit. Bio-optical systems lose a lot of power through radiation (light) emission and data processing. These systems are similar to those described aboveFor reference, it may run for 5 days with one charge. It is to be noted that it is preferable that,continuous optical monitoring is not really performed. PPG (optical module) typically consumes 2.3mA and the power consumption depends on the number of sensors. An 8-channel optical system and inertia measurement are used, the power consumption is 19.4mA, and the working time is 5-10 hours.

In one embodiment, the control device includes an array of pressure sensors arranged in a continuous flexible structure within an arc of a belt or other structure for maintaining the array of sensors in close proximity to a body limb. The pressure sensor array is comprised of a material that is substantially free of any of glass, rigid transparent polymer, stainless steel, or light emitting or detecting means. These structures are not included in the individual pressure sensors of the array, although these components may be included in the control device, or may be included in any component of the data storage, data processing, logic circuitry, memory or communication components of the control device.

Example 3-quantitative analysis of pressure sensor data and three-point analysis of body position.

In a preferred embodiment, the pressure sensor data is quantitative. For example, a closed fist (characterized by muscle tension) differs in number from an open fist (characterized by fingers bending and no muscle tension) and still from other hand positions, such as an open palm. Similarly, as the local pressure of the tendons controlling the movement of the finger increases, the variation and dispersion of the finger's curvature can also be addressed. The data from the array of pressure sensors may include three separate data points including a first initial or rest position of the user's limb at which no motion occurs, a transitional phase of motion, such as intentional movement of the user's wrist, hand, or single finger, occurring away from the first position, and then a second position differing in number from the first position as a result of the transitional phase motion and resulting in the second position. Additional actions different from the first location and/or the second location and having a subsequent transition phase may also be detected and quantified.

The electrical system (EMG) for posture detection relies on signal detection during muscle contraction. This requires that the electrodes be close to the contracting muscles driving the hand/finger movement and that the muscles have to contract actively. This means that only states characterized by active contraction can be measured. Passive hand positions have no significant electrical signal, such as an extended finger or a loose fist. To detect a fist, the hand must be clenched. This limitation is only used to detect movement/translation and only long time states can be detected during contraction, which can quickly lead to muscle fatigue. EMG signals are characterized by high levels of noise resulting from EMF interference, motion artifacts, small (μ V) input signals, bioimpedance changes, and biointerface changes. Signal amplification and de-noising of these signals for specific gesture detection is power consuming (typically 2-4mA per channel) and computationally intensive.

The optical system (operating using similar equipment and principles as PPG, as described in 2016/0091980, a photoplethysmogram) relies on a light source to penetrate and reflect human tissue and tracks the intensity of this reflected light through a detector. The main limitations of this system are power consumption from continuous radiation emission (typically 2mA per emitter/detector pair) and susceptibility to noise. The latter is particularly troublesome. Although changes in hand position can be detected (no active muscle contraction is required), the baseline offset of the noise (recorded in section [0039] at 2016/0091980) and light signal is such that only transitions can be reliably detected. The 2016/0091980 publication records motion detection in fig. 9B, and the detection of motion is particularly noted in the description of paragraph [0040 ]. For optical systems that disable absolute position detection, a key difference in pose/position detection is that the detected light has no direct relationship to the tendon position. That is, no specific light flux necessarily indicates tendon contraction/relaxation. The signal input into the detector is affected by ambient light, skin reflections, pressure between the optical system and the skin, and tendon/muscle/bone position. The changes caused by the noise source typically exceed the changes in tendon position by an order of magnitude or more. In practice, this makes absolute position detection without posture change (e.g., long time state of extended index finger) difficult, if not impossible.

The basic biological parameter measured by pressure or optical systems is tendon displacement due to contraction and relaxation. In the case of a pressure-based system, this tendon displacement must exert pressure on the band that is tight around the wrist. In the case of known band tensions, the pressure is directly related to the displacement of the tendon. So that a large displacement (e.g., fist) will generate a pressure of about 10mmHg, while a small displacement (e.g., little finger straightening) will result in a pressure of about 1 mmHg. In a well-calibrated system, the quantitative pressure at a given location is of well-defined importance to the underlying tissue. In this way, even without motion, the circumferential pressure distribution of the wrist can be used to accurately access tendon/muscle/bone states and determine hand position/posture states.

Example 4-learning mode for pressure-based gesture analysis

The control device or companion device preferably includes a memory device to store pressure sensor data that generates a unique pressure profile that includes various aspects of the user-initiated action. The pressure profile characteristics are quantified as: 1) the value of each individual pressure sensor, 2) the change in the individual value of each or all of the individual pressure sensors, 3) a plurality of individual or combined values, an overall quantified score resulting from one or more metrics resulting from the input of the plurality of pressure sensors, 4) any of the above at discrete points in time, 5) any of the above at a plurality of points in time, including the above rate of change and the change in the above rate of change, all depending on the user's location or motion.

The control device may also compare the unique pressure profile produced by the pressure array to one or more stored values or profiles comprising the unique pressure profile to determine whether the individual user action or set of actions corresponds to the stored unique pressure profile, and the comparison may generate any of a signal, a new score based on the comparison, a determination of the action corresponding to a command generated by the control device, or an absence of such a determination. In one embodiment, the control device compares the unique pressure profile to a stored range of predetermined pressure profiles and associates the sensed pressure profile with the most similar stored pressure profile and generates a command that is transmitted to the companion device. In another embodiment, the companion device has a visual or mechanical option displayed on the companion device to ask the user whether the interpreted control or command signals generated by the control device are correct.

The control means may be adapted to filter noise caused by user action that is not unique to the generation, determination or analysis of the unique pressure profile or score. The noise may comprise random movements of the user, voluntary physiological functions (e.g. breathing, heart rate) or any extraneous or partial background signal that is different from the intended motion of the user. The filtering function may be based on a pressure range, the presence of a known physiological function established as part of a baseline measurement, or any other factor that distinguishes noise from an intended user-generated signal resulting from data generated by the pressure sensor.

In a preferred embodiment, either the control device or the configuration device contains data storage and logic circuits or functions to allow one or both devices to operate in a learning or teaching mode, wherein user actions are responsive to instructions from either device instructing the user to perform a particular action or gesture that is subsequently intentionally performed by the user to provide the user with a stored unique pressure profile, respectively, and to compare with subsequent gestures performed in a standard mode of operation. The user may also identify a particular gesture associated with a particular command based on the user's indication, such that the user teaches the control device that a particular output generated by the control device command is associated with a particular action or set of actions or a particular gesture.

In some embodiments, the detachable and portable control device comprises at least six pressure sensors embedded in an arc of a flexible and wearable device having an angle of at least 120 degrees, the flexible and wearable device positioning the pressure sensors on a circumference of a skin surface of a body part of a user, wherein a first quantitative measured pressure value or set of values is associated with a first location of a limb of the user, a quantitative interim pressure value or set of values is associated with an action of two or more physical structures selected from bone, ligament, tendon, epidermal layer, muscle and interstitial tissue, and a second quantitative measured pressure value or set of values is associated with a second location of the limb of the user. The control device is operatively connected to a data processor containing memory and logic capable of determining the user's actions is associated with commands to be communicated with the control device or companion device.

In an embodiment, the plurality of pressure sensors is located in a control device comprising an array of pressure sensors embedded in the strip, at least 8, at least 16, at least 24, at least 32 or at least 64 individual sensors being arranged circumferentially around the strip. The sensor array is operatively connected to a data processor containing memory and logic capable of determining an action detected by the circumferential pressure array is associated with a command to be communicated with a control device or companion device. In some embodiments, the companion device is also worn by the user and is operatively connected to the control device by wireless transmission or by a mechanical connection such that the companion device is controlled by the portable electronic device which in turn is controlled by the pressure signal detected by the user's motion and detected by the control device.

In some embodiments, the control device includes an array of pressure sensors individually and independently located in a structure that holds the companion device at the skin surface proximate to one or more tendons, ligaments, bones, or muscles located in the limb. For example, the companion device may be a smart watch and the control device may include a sensor array incorporated into a band comprised of a plurality of sensing zones spaced apart from the companion device and detecting pressure changes at a pressure sensor interface proximate to a zone of the smart watch and across an adjacent surface of the user's skin.

The present invention also includes a method of distinguishing multiple individual gestures of a user using the control device described herein so that unique gestures can be translated into commands for controlling a companion device. A method of executing a command based on a user action may include the step of determining that a gesture of the user conforms to a predetermined pressure profile detected by the sensor array. The method can comprise the following steps: detecting a signal comprising a pressure sensor value, wherein the value is a quantitative value comparing the unique pressure profile of the first rest position with a transition pressure value reflecting a user action; detecting a change in pressure associated with a user action; and determining that the signal corresponds to a specific command for controlling the control device or the companion device.

With general reference to the figures, the present invention relates to a pressure-based motion detection apparatus for detecting a specific body motion and converting the body motion into an output. The human body motion is converted into a unique pressure signal and converted into a control command. In some embodiments, the physical action includes a specific movement of the hand, and typically includes a recognized orientation of the forearm/wrist, palm and fingers that produces a unique pressure signature. The unique pressure signal generated by the human body action (e.g., recognized gesture) is processed as a measured combination of sensor inputs based on measured pressure, measured pressure changes, or measured pressure changes over time. In this manner, body motion is quantified based on unique pressure characteristics and converted into commands and control instructions that provide instructions to a control device or a companion device operatively connected to the control device.

Because the pressure sensors described herein are inexpensive to manufacture and provide extremely high pressure sensing performance as described in example 2, multiple sensors can be incorporated into a control device that can detect discrete actions separately and simultaneously from a large number of discrete pressure measurements resulting from the action location of physiological structures under the skin, providing the ability for a large number of individual pressure monitoring data inputs. The large amount of quantitative data input allows the assembly of extremely sensitive and selective unique pressure profiles that are discrete and distinct from the specific movements of the body. Thus, the array of pressure sensors may detect user-specific body movements, including retraction, extension, expansion, contraction, rotation, and almost any other movement of the user, voluntarily or involuntarily, voluntarily or voluntarily, such as hand movements in preselected gestures, or passively or automatically, such as breathing, pulse, or blood pressure.

In particular, with respect to the movements of the hand and wrist detected by pressure changes resulting from the unique configuration of the forearm/wrist, hand and fingers, tendon measurements, movements of individual components or sets are determined and may be characterized as flexion, extension, abduction and adduction, with the combination of the various metrics as a score, with a weighted distribution of the change over time or rate of change for each movement, position, movement or position. Flexion is generally defined as the motion of the hand with the palm moving toward the body side, and extension is generally described as the motion of the hand with the palm moving away from the body side. Abduction is generally defined as the movement of the hand with the palm moving towards the side of the little finger and adduction is generally defined as the movement of the hand with the palm moving towards the side of the thumb. Finger flexion is defined as movement of the finger towards the wrist, while extension is defined as straightening of the finger.

With respect to any point on the body, motion can be characterized as absolute or relative motion along any X, Y or Z axis, any rotation about axis θ, γ, or φ, the change in position of any point or set of points over time, and the change in rate of change of motion or rotation over time. In addition to pressure-based sensor arrays, additional components such as accelerometers, gyroscopes, and magnetometers may be incorporated into the control device and similar change measurements, rate of change measurements, and rate of change measurements may be made over time. In a preferred embodiment, three accelerometers, three gyroscopes and three magnetometers are combined with at least eight ion-based pressure sensors for generating unique pressure profiles. Combinations of absolute or relative motion may be combined with changes in absolute or relative pressure to create highly selective and specific motion and pressure-based commands that are translated into specific outputs, such as commands or control functions for a companion device.

Most movements of the elbow and shoulder, in some cases, the user's actions have no significant pressure effect on the hand and wrist, such as displacement along the X, Y or Z axis, including some degree of linear or arcuate movement or rotation, and we can analyze as independent measures. Some user actions, such as internal and external rotation of the hand, palm up or down, because both pressure and rotation changes can be detected at the skin surface using the pressure array. The combination of gestures, which combines the movement of the limb and the rotation, results in a selection of more different position states and characteristic movements, which can be analyzed and scored. For example, the thumb-up position, thumb-side position, and thumb-down position are all different position states and rotations, wherein a combination of a pressure sensor and a motion sensor as described herein may generate a pressure signal and a motion signal. Repeated gestures (e.g., tapping the index finger and thumb together twice) can provide greater accuracy for interpretation of any action or gesture and increase the signal component of the pressure score, as distinguished from other common actions properly characterized as noise.

In operation, the end result is that pressure changes sensed around an area of the user's body are translated into commands for a companion device such as an image or video display controller, a gaming machine controller, or a computer user interface. As known to the user, the body motion is instantaneously translated into recognizable commands for any companion device. Thus, when a user performs a gesture, the composition of the underlying tissue imposes a unique and characteristic pressure profile on the sensor array, which, possibly in combination with other motion sensing components, translates the gesture into a specific command that is immediately executed in the companion device.

The control device or companion device includes logic for comparing any value, set of values or scoring a stored reference value or comparison value, optionally considered as a threshold value, to determine that the body motion corresponds to a stored body motion parameter paired with a particular control command. The control means generates a predetermined command if it is determined that the body motion matches the stored value, distribution or score. For example, any forearm/wrist, hand or finger gesture may rely on the generation of a particular pressure detection feature, and the set of motions of the forearm, wrist, hand or finger or any discrete physiological structure may be separated so that the motions of the hand in any subset of fingers may be distinguished from the motions in a different subset of fingers and similar or different gestures. In this manner, any individual physiological component, the position or motion of which can be assigned as an empty set or control value, the individual motion of any physiological structure being indicative of the output of a particular control or commanded control device.

Equivalent distinctions in quantitative measurements may also be based on absolute or relative time characteristics, such that the absolute or relative value of any pressure reading may be expressed or analyzed as a function of time, either together with or separately from any other measurement. In this manner, the time interval T1 may be determined for any particular pressure sensor reading, profile, or fraction such that the difference in the time interval from time point T12 to time point T2 determines only the difference in the output from the control device. In other words, slower actions are distinguished from faster actions, and the difference between the time required to make any action or change in position may itself lead to degradation of command control.

As mentioned above, the logic of the control device may be adapted to filter noise signals generated by external pressure, such as pulses, including heart rate, respiration, blood flow or other autonomous functions over time. To this end, the control device may accept input from reference sensors specifically positioned to account for noise separated from the main sensor array signal. The control device may also include logic for sending the pressure sensor array output to a companion device or elsewhere so that the pressure sensor array data can be further processed to generate command and control signals as described herein. In such embodiments, the companion device includes a corresponding receiving unit such that the companion device and the control device wirelessly communicate with each other.

In the method of the invention, the pressure sensor array first measures a starting, rest or raw first position before measuring any signal from any accelerometer, gyroscope or magnetometer based on quantitative pressure sensor data. Subsequently, changes in the pressure sensor array values relative to the first position caused by the user's body movements are determined and quantified, optionally in combination with input from any accelerometer, gyroscope, or magnetometer. Subsequently, after the transitional movement phase, the pressure sensor data or the change in the pressure sensor data is determined for a different second position, which is different from the first position. The difference between the first position, the transition phase or the second position is measured and associated with the position or the motion of the user's body. For example, if the body motion is a hand motion comprising motion of the entire hand or motion of a portion of the hand (e.g., motion of one or more fingers relative to another portion of the hand), a difference in position of the hand is determined based on a difference in pressure sensor values, or a comparison score is determined to indicate a predetermined hand motion and sent by the control device 1 to a companion device controlled by the hand motion.

The method may include a termination or verification step in which a display or sensory input is provided to the user to confirm or deny that the gesture performed by the user generated the correct command or control instruction. Once the user makes a selection, the gesture will be ignored and the logic circuit will discard the measurement data from the first location through the transition phase to the second location. If the gesture is confirmed, a command or control is executed.

Typically, the signals from the accelerometer, gyroscope and magnetometer are independent of the signals generated by the pressure array, but can be analyzed along with the pressure array data to generate an output for the control device. In some cases, the pressure sensor and the position sensor may be linked together and scored. For example, user motion, primarily elbow and shoulder, results in less signal in the pressure sensor and more signal in the motion sensor, while hand motion results in more signal in the pressure sensor and less signal in the motion sensor. One important exception is the internal/external rotation of the hand (up or down), caused by the rotation of the elbow, where pressure changes on a pressure array disposed on the wrist may produce a dominant signal that is greater than the motion-based signal.

Although in the above described embodiments the detected body motion is used to control the companion device, in some embodiments the body motion may be used to control the control device itself, for example to power the device on or off, or to initiate communication with the companion device.

Referring specifically to FIG. 1, six individual metrics (G0-G5) represent the individual elements of the quantitative pressure sensing data distribution of the control device of the present invention. The combination of metrics based on the individual magnitudes of the individual metrics G0-G5 yields a unique quantitative gesture detection score based on a particular individual pressure value input from the sensor array, and may be reflected as a region according to any combination of the contribution magnitudes of the individual metrics G0-G5. In fig. 1, the combination of the individual metrics is shown by the shaded area arranged around the center point of the graph. Those skilled in the art will appreciate that this graph and the particular representation of a single metric G0-G5 are chosen for convenience only, and that any similar graph may convey the same concept of a combined score based on multiple measurements of various metrics of a human. Furthermore, the particular number of individual metrics shown in FIG. 1 is merely representative. A different number of individual metrics may be selected, regardless of the total number of sensors. Thus, the number of individual metrics may be less than, equal to, or greater than the total number of sensors.

In any configuration, the input of individual metrics may be used to generate quantitative scores based on individual or collective contributions of the metrics. Furthermore, the contributions of the individual metrics may be weighted such that one or more metrics contribute more or less to the final score. Still further, each individual metric may be analyzed over time to determine the change in metric over time, including the rate of change of any individual metric or set of metrics over time. The example of FIG. 1 shows six separate metrics G0-G5 from pressure sensor data. In addition, separate sensors other than pressure data may be integrated into the profile to change the profile shape in the X-Y plane, or other features (e.g., Z-direction out of the plane of the page) may be provided to further enhance the work profile. As described above, in any metric change, including their absolute or relative values, changes in value, rates of change in value, changes in rates of change, the distributions may be static or dynamic and associated with gestures or gesture combinations.

Any absolute or relative measure of motion, range or pattern of motion may be analyzed and used to generate a quantitative score that may reflect either the combined pressure measurement alone or compared to a previous single or combined set of actions. The individual or collective action scores may be associated with types of actions performed by the user that are characteristic of functions to be performed by the user or characteristics of controls or commands to be controlled by the user to the companion device.

Typically, a single metric will not match one-to-one with a specific number of sensors or a specific anatomical structure (e.g. tendon), since each metric may be a contribution from one or more pressure sensors, and a combined metric reflects a combined pressure distribution with combined contributions from multiple anatomical structures. The pressure distribution scores corresponding to a particular single or discrete motion, range of motion, or combination or subset of motions may be identified and used to establish a motion pattern or deviation from a pre-existing pattern.

With particular reference to FIG. 2, a cross-section of a human wrist shows the location and orientation of various physiological structures including the 14 tendons T1-T14, the radius and ulna, and other internal anatomical structures such as cartilage, interstitial tissue, nerves, and joints. Placement of multiple individual pressure sensors conforms closely to the outer skin layer of the wrist, creating unique quantitative pressure profiles based on the different configurations and motions of the forearm, wrist, hand and fingers. Due to the large number of anatomical structures and their unique location and orientation within the range of human user motion, skin surface pressure variations can vary greatly, enabling accurate measurements by placing a sufficient number of pressure sensors with sufficient selectivity and sensitivity to detect these variations.

As described above, the pressure sensor array of the present invention can be incorporated into a band-like structure around a limb (e.g., a wrist) to place the sensors in a position that can detect basic physiological structure movements reflected by changes in skin surface pressure. As shown in fig. 2, tendons, bones, and other physiological structures are not uniformly distributed in the horizontal or vertical direction, as described in more detail below. Similarly, as a person moves their arms up or down, the position and shape of the various structures change, and thus the position of the strap around the wrist may be highly independent to the user. The selection and positioning of the pressure sensors in the array can be tailored to the specific location of each tendon, bone, or other structure to specifically detect positional differences in certain movements.

It is evident from the relative positions of the sensors that the pressure reading from a single sensor will reflect the physiological structure in the vicinity of the point of contact between the sensor and the area of skin in conforming engagement with the sensor. Since the physiological structures in the wrist are asymmetrically oriented with respect to the outer surface of the skin, each sensor must reflect pressure changes from different combinations of tendons, bones, muscles, and any other structure that creates a pressure differential across the sensor. Thus, the pressure data delivered to a single sensor may be unique to its location around the outside of the circumference of the wrist. Thus, the band in which the sensors are arranged is designed to be placed around the wrist in the same direction each time it is used alone. For example, the band may have a tensioning element that maintains the end in consistent engagement with the outer surface of the skin for positioning orientation of the sensor array such that each sensor repeatedly senses the same combination of physiological structures, such that the sensor array data profile produces similar data upon repeated use, and thus the data sensor profile may be stored in a data storage element for comparison to each user.

With particular reference to fig. 3A and 3B, the control device 10 includes a plurality of sensors, numbered 1-8 in fig. 3A and 1-16 in fig. 3B, and a housing or case 11 for housing data storage, power supply and logic circuitry to detect and analyze the unique pressure signatures described herein. Referring to FIG. 3A, sensor array 20 includes a collection of individual sensors 1-8 that conform closely to the outer surface of the user's limb. As mentioned above, the sensor array is preferably arranged in a belt with tensioning elements for maintaining contact between the sensor array and the outer surface of the skin to detect pressure variations at each point around the array to a maximum extent. As with the wristband, the design of the strap and tensioning element repeatedly places the sensor array in the same position relative to the user's underlying physiological structure so that each subsequent use can reliably compare the pressure sensor data distribution of a previous use or calibration to accurately interpret the user's posture.

In the embodiment of figure 3A, the sensors surround the entire circumference of the limb, except for the portion occupied by the housing 11, so assuming that the housing is approximately the same size as any single sensor, the sensors surround approximately 320 ° of the limb. The embodiment of fig. 3A shows an identifiable example of a sensor array 20 arranged in a watchband, wherein the functional parts of the watch elements are arranged in a housing 11, which is attached or integrated into the watchband, as in a conventional watch or digital watch. Preferably, the sensor surrounds at least 180 °, 200 °, 220 °, 240 °, 260 °, 280 °, 300 °, 320 °, 340 ° and 360 °. The number of sensors may be larger, equal to or smaller than the number of tendons. Assuming there are 14 tendons as shown in fig. 2, each of the 8 sensors in the fig. 3A embodiment receives pressure data from more than one tendon as well as the underlying bone, ligaments, joints, subdural tissue and vascular system. Furthermore, although the individual sensors in the arrays 20 in fig. 3A and 3B are substantially equal in size and equally spaced around the limb, other embodiments may include sensors of different sizes and whose spacing around the limb is off-center to take advantage of the unique pressure sensors that may result from the basic physiology of different points around the wrist.

With particular reference to the embodiment of fig. 3B, the sensor array 20 includes 16 individual sensors 1-16 that surround the entire limb of the user, such that the housing 11 is separated from the skin layer and disposed along the surface of one or more components of the sensor array 20. Thus, the coverage of the sensor array around the circumference of the limb is 360 °. As described above, in the present embodiment, the number of sensors 1-16 exceeds 14 individual tendons, and since more than one sensor can be designated to interpret sensor data input from a single tendon, the selectivity and specificity of the control device 10 increases. This configuration increases the sensitivity and specificity of the array, and improves the accuracy of the array in interpreting the individual pressure-based distributions. Furthermore, as the number of individual sensors T1-Tn in sensor array 20 increases, smaller changes in the user's motion may be detected and analyzed to produce a more accurate pressure signature distribution, and finer motions and gestures may be detected.

Referring specifically to fig. 4, normal motion of a human limb may be detected by a simple change in pressure reflected in the output of the sensor array and associated with a plurality of actions or gestures based on apparent and recognized changes in the array input. With particular regard to arm, hand and wrist motion, detected pressure changes due to such individual and collective motion of the arm, hand and wrist may be detected and associated with the respective motion components. The individual components of motion may be characterized individually or collectively as combinations of flexion, extension, abduction and adduction, and each measurement, and may optionally be described as a score with a weighted distribution of the change in time or rate of change over time for each motion, position, action or position. Each different combination of arm, hand and wrist motion produces recognizable ground pressure data profiles for a particular motion or gesture, which can be used to generate an output signal to a companion device, or stored for further analysis, comparison with a future set of motions or gestures, or used as part of a calibration procedure.

Referring again to fig. 4, flexion is generally defined as the motion of the hand with the palm moving toward the body side, and extension is generally described as the motion of the hand with the palm moving away from the body side. Abduction is generally defined as the movement of the hand with the palm moving towards the side of the little finger and adduction is generally defined as the movement of the hand with the palm moving towards the side of the thumb. Finger flexion is defined as movement of the finger towards the wrist, while extension is defined as straightening of the finger. "N" indicates that the orientation of the anatomical structure is in a neutral state or position. The level 1 pressure profile is most easily detected, including arm movement, palm flexion or extension, and hand opening or closing as detected by finger motion. Level 2 motion changes are more difficult to detect, including adduction or abduction of the palm. The level 3 pressure signal is more difficult to detect and needs to distinguish between the individual movements of the finger.

As indicated, measuring the difference between thumb and forefinger, either together with or separately from other fingers or anatomical structures, allows the pressure data profile to differentiate between forming an "okay" signal, indicating the forefinger, giving a "good" or "thumb" gesture, extending the forefinger, or forming a "gun" with fingers and palm, and opening or closing the palm. As described above, the pressure data profile generated by each set of actions may be converted to an output signal, stored or used as part of an analysis or calibration process, wherein the pressure data profile generated by a series of actions is stored and used for comparison with future actions resulting in the output signal and the determined configuration of the companion device. For example, a user wishing to issue a "thumbs up" gesture to activate the computer terminal may be instructed to repeatedly execute the structure so that the pressure sensor array measures a set of action features to generate a pressure data profile that is specific to the user and specific to the measured pressure differential caused by the selected action of the user's wrist physiological structure. Once a sufficient baseline of data is collected, the device is functionally calibrated to recognize and convert future characteristic actions or gestures of the user into output signals.

Referring to FIG. 5, a simple list of commands shows how a single function can be assigned to a companion device such as a game controller, computer mouse, or in a software program (e.g., a game console), using a single action of the wrist, hand, and fingers) To generate a single command. Taking the computer mouse in the middle column of fig. 5 as an example, the open hand gives an instruction to "drag and drop" the highlighted field of the quality controller. Closing the hand commands the drag field, while flexion of the hand represents "undo" or "rollback" and extension of the hand represents "forward". Forming a circle with the thumb and forefinger will result in a "left click" output signal, command or control, while forming a circle with the thumb and ring finger will result in a "right click". Abduction or adduction generates control commands for turning pages up or down, respectively.

Referring to fig. 6, fig. 6 is a graph showing representative changes in physiological interface pressure across the right wrist and as indicated by fig. 2 and 3A-3B for 14 individual tendons labeled T1-T14. Although the numbering and orientation are arbitrary, in the pressure signal, motion detection and posture command schedule depicted in fig. 6 and in the exemplary device of fig. 3B, the pressure sensor measures and quantifies changes in the tendons 1-6 of the wrist/volar flexor forearm and the extensor wrist/dorsal forearm tendon using five pressure change categories indicated by arrows. A wide range of detected pressure changes for each of the 14 tendons of a particular user action is measured by these criteria and translated into a posture. Referring to the various boxes of fig. 6, the blank fields indicate no change in pressure, a single downward arrow indicates a small decrease in pressure value, two downward arrows indicate a large decrease in pressure value, a single upward arrow indicates a small increase in pressure value, and two upward arrows indicate a large increase in pressure value. By associating a specific change with the pressure at each tendon, a pressure data distribution consisting of a plurality of pressure change data is detected and analyzed for pattern recognition and association with different gestures. The left axis contains at least 11 examples of individual direction or command-control functions that can be assigned to a particular combination of measured changes in trans-tendon pressure.

Thus, for the hand-open posture, the pressure sensor array 20 as shown in fig. 3B detects a large pressure increase of the tendons 1, detects a small pressure increase of the tendons 2-4, detects a large pressure increase of the tendons 5-10, detects a small pressure decrease of the tendons 11-13, and detects a large pressure increase of the tendons 14. For the remaining poses shown in the left column, similar pressure changes may be associated on 14 tendons. Although the number and orientation are arbitrary, in the pressure signal, motion detection and gesture command list shown in fig. 6 and in the exemplary apparatus of fig. 3B, the pressure sensor generates a particular pressure data profile that is further processed and can be converted into a signal that is output to the companion device. As described above, because the individual physiology and selected actions of the user are unique to the individual, unique pressure sensor data can be used to generate a pressure data distribution that is unique to the individual's actions or gestures. Thus, in practice, the same person making the same "okay" gesture will produce a pressure data profile that is significantly different from another person. These differences can be used to identify individuals for security purposes and to activate unique identifiers such as logins, passwords, locks, and any other mechanism that requires personal identification. Importantly, unlike conventional input devices such as keyboards or mice, gestures are only detected from an array of sensors affixed to the user's limb, so no physical contact is required between the user and the input device. When integrated into a watch, a single gesture may also operate any function of the watch, including sending or receiving messages, making or receiving calls, activating instructions for the electronic device and the user interface, and generating output signals to control various electronic devices by simple gestures according to a user's unique personalized pressure data profile.

Although the disclosed examples have been fully described with reference to the accompanying drawings, it is to be noted that various changes and modifications will become apparent to those skilled in the art. Such changes and modifications are to be understood as being included within the scope of the disclosed examples as defined by the appended claims.