Medical cart anti-collision detection method, device and system

文档序号:1446339 发布日期:2020-02-18 浏览:4次 中文

阅读说明:本技术 医疗推车防碰撞检测方法、装置和系统 (Medical cart anti-collision detection method, device and system ) 是由 郭永安 韩镇宇 刘云 朱洪波 杨龙祥 于 2019-10-21 设计创作,主要内容包括:本发明提出了一种医疗推车防碰撞检测方法、装置和系统。方法包括:S1、采集医疗推车四周的障碍物信息,根据PCA算法将三维障碍物坐标信息降维成低维坐标信息;S2、利用OBB有向包围盒目标检测方法生成障碍物的边界框;S3、计算障碍物边界框与医疗推车的边界框之间的距离,并根据距离进行碰撞预警和告警;S4、通过TCP/IP协议将碰撞预警、碰撞告警信息上传至行车记录管理平台。本方法通过物联网、云计算、机器学习等信息技术,开发医疗推车的行车记录管理平台,利用OBB有向包围盒碰撞检测技术,做到提前预警、碰撞记录等,实现医疗推车安全行进与统一管理。(The invention provides an anti-collision detection method, device and system for a medical cart. The method comprises the following steps: s1, acquiring barrier information around the medical cart, and reducing the three-dimensional barrier coordinate information into low-dimensional coordinate information according to a PCA algorithm; s2, generating a boundary frame of the obstacle by using an OBB directed bounding box target detection method; s3, calculating the distance between the boundary frame of the barrier and the boundary frame of the medical cart, and performing collision early warning and alarm according to the distance; and S4, uploading collision early warning and collision warning information to the driving record management platform through a TCP/IP protocol. According to the method, a driving record management platform of the medical cart is developed through information technologies such as Internet of things, cloud computing and machine learning, early warning, collision record and the like are achieved by utilizing an OBB directed bounding box collision detection technology, and safe advancing and unified management of the medical cart is achieved.)

1. A medical cart anti-collision detection method is characterized by comprising the following steps:

s1, acquiring barrier information around the medical cart, and reducing the three-dimensional barrier coordinate information into low-dimensional coordinate information according to a PCA algorithm;

s2, generating a boundary frame of the obstacle by using an OBB directed bounding box target detection method;

s3, calculating the distance between the boundary frame of the barrier and the boundary frame of the medical cart, and performing collision early warning and alarm according to the distance;

and S4, uploading collision early warning and collision warning information to the driving record management platform through a TCP/IP protocol.

2. The medical cart anti-collision detection method according to claim 1, wherein the step S1 includes:

s1-1, collecting three-dimensional random vector X ═ of obstacles around the medical cart (X ═1,X2,X3)TN samples xi=(Xi1,Xi2,Xi3)TI 1,2,., n, n > 3, constructing a sample array, and carrying out the following operation on sample array elementsAnd (3) normalization transformation:

Figure FDA0002240731210000011

wherein

Figure FDA0002240731210000012

s1-2, solving a correlation coefficient matrix for the normalized matrix W:

Figure FDA0002240731210000013

S1-3, solving characteristic equation | R-lambada I of sample correlation matrix R3Get 3 characteristic roots if 0:

push button

Figure FDA0002240731210000015

S1-4, converting the normalized index variable into a main component:

Figure FDA0002240731210000017

u1 is called the first principal component, U2 is called the second principal component, and Up is called the pth principal component;

s1-5, carrying out weighted summation on the m main components to obtain a final evaluation value U*=(U1,...,Um)。

3. The medical cart anti-collision detection method according to claim 1, wherein the step S2 includes:

s2-1, randomly selecting n two-dimensional coordinate points { (x1, y1), (x2, y2),. } and (xn, yn) } as known points on the boundary of the detection object;

s2-2, combining the data according to the coordinate types: x ═ x1, x 2.., xn }, y ═ y1, y 2.., yn };

s2-3, respectively calculating the average values averX and averY of the set x and the set y, and constructing a covariance matrix:

Figure FDA0002240731210000021

s2-4, solving eigenvalues and eigenvectors of the OBB bounding box according to the covariance matrix and the Jacobian algorithm, and selecting the largest eigenvalue as the direction of the OBB bounding box;

s2-5, establishing a new coordinate system according to the obtained characteristic vectors, putting the original data back under the new coordinate system, and solving the length, width and central point of the OBB bounding box;

s2-6, length, width, center point, and OBB bounding box orientation constitute the bounding box of the obstacle.

4. The medical cart anti-collision detection method according to claim 1, wherein the step S3 includes:

s3-1, establishing a model according to the coordinate axis of the OBB bounding box and the coordinate axis of the bounding box of the medical cart, wherein the model describes the rotation angle of the coordinate axis of the OBB bounding box relative to the coordinate axis of the bounding box of the medical cart;

s3-2, calculating the distance between the boundary frame of the medical cart and the boundary frame of the obstacle, judging that a collision danger exists when the distance is smaller than a first threshold value, and triggering early warning prompt; and when the distance is smaller than or equal to a second threshold value, judging that collision occurs, and triggering an alarm prompt.

5. The medical cart anti-collision detection method according to claim 4, wherein the step S3 further includes: the medical trolley is sensed by the vibration sensor to vibrate, when the distance between the boundary frame of the medical trolley and the boundary frame of the obstacle is smaller than or equal to a second threshold value and the vibration sensor senses vibration, collision is judged, and an alarm prompt is triggered.

6. The utility model provides a medical shallow anticollision detection device which characterized in that includes: the medical trolley comprises a video acquisition module, a vibration sensor, an information display module, an information storage module, an alarm module and a control module, wherein the video acquisition module is distributed around the medical trolley, acquires video image information of obstacles and sends the video image information to the control module; the vibration sensor converts the sensed vibration into an electric signal and sends the electric signal to the control module; the control module processes the acquired barrier video image to obtain three-dimensional coordinate information of the barrier, and performs dimension reduction processing on the three-dimensional coordinate information to obtain two-dimensional coordinate data of the barrier; calculating the distance between the boundary frame of the barrier and the boundary frame of the medical trolley, judging collision by combining vibration signals transmitted by the vibration sensor, and controlling the display and alarm work of the information display module and the alarm module; the information display module provides an interactive interface with a medical cart user and can display various information; the information storage module is used for storing data, including collision information, collected video information and medical cart use records; the warning module triggers a warning or warning event under the control of the control module when a warning or warning condition is reached.

7. A medical cart anti-collision detection system is characterized by comprising the medical cart anti-collision detection device and a driving record management platform, wherein the medical cart anti-collision detection device is connected with the driving record management platform through a TCP/IP protocol, and when a medical cart enters a wireless local area network coverage range, data stored locally by the detection device can be uploaded to the driving record management platform through touch screen operation of an information display module of the medical cart; the driving record management platform also provides the functions of user management, personal center, alarm recording, health promotion and education and problem consultation.

Technical Field

The invention belongs to the technical field of intelligent medical treatment, and particularly relates to an anti-collision detection method, device and system for a medical trolley.

Background

Wisdom medical treatment is moving towards the life of common people, and wisdom medical treatment utilizes advanced internet of things through creating regional medical information platform, realizes the interdynamic between patient and medical personnel, medical unit, the medical equipment, reaches informationization gradually. The frequency of use of medical carts throughout a medical work session is very high in the daily routine of medical workers. The medical cart is a medical auxiliary appliance commonly used in hospitals, and comprises various types, including: transfer vehicle (stretcher, ambulance): pushing the patient from the emergency ambulance to the operating room, or from the hospital bed to the operating room; a medical vehicle; an anesthesia trolley; a transfusion vehicle; the sewage vehicle (applicable range is hospital supply room, ward, operating room and laundry); a medicine delivery vehicle and the like. Due to the urgent work of hospitals, patients need to be treated quickly and timely, and the collision accident of the medical cart is easy to happen. When the medical cart is collided, public facilities such as an elevator and the like are often damaged. If the transfer trolley collides to generate vibration in the process of transferring from the emergency ambulance to the operating room, the secondary injury of the patient can be caused; the medical instrument can be damaged when the medical vehicle collides with the pedestrian or other transfer vehicle in the hospital passageway or the elevator opening. If the situation occurs, in the traditional processing method, the hospital security department searches for the collision of the medical cart through video monitoring, and therefore the workload is large and the efficiency is low. Therefore, it is an urgent problem in the field of intelligent medical treatment to research an anti-collision method for medical carts.

Disclosure of Invention

The purpose of the invention is as follows: in order to solve the problems in the prior art, the invention provides an anti-collision detection method, device and system for a medical cart, which can effectively realize safe advancing and unified management of the medical cart.

The technical scheme is as follows: according to a first aspect of the present invention, there is provided a medical cart collision avoidance detection method comprising:

s1, acquiring barrier information around the medical cart, and reducing the three-dimensional barrier coordinate information into low-dimensional coordinate information according to a PCA algorithm;

s2, generating a boundary frame of the obstacle by using an OBB directed bounding box target detection method;

s3, calculating the distance between the boundary frame of the barrier and the boundary frame of the medical cart, and performing collision early warning and alarm according to the distance;

and S4, uploading collision early warning and collision warning information to the driving record management platform through a TCP/IP protocol.

According to a second aspect of the present invention, there is provided a medical cart collision avoidance detection apparatus comprising: the medical trolley comprises a video acquisition module, a vibration sensor, an information display module, an information storage module, an alarm module and a control module, wherein the video acquisition module is distributed around the medical trolley, acquires video image information of obstacles and sends the video image information to the control module; the vibration sensor converts the sensed vibration into an electric signal and sends the electric signal to the control module; the control module processes the acquired video image to obtain three-dimensional coordinate information of the barrier, and performs dimension reduction processing on the three-dimensional coordinate to obtain two-dimensional coordinate data of the barrier; calculating the distance between the boundary frame of the barrier and the boundary frame of the medical trolley, judging collision by combining vibration signals transmitted by the vibration sensor, and controlling the display and alarm work of the information display module and the alarm module; the information display module provides an interactive interface with a medical cart user and can display various information; the information storage module is used for storing data, including collision information, collected video information and medical cart use records; the warning module triggers a warning or warning event under the control of the control module when a warning or warning condition is reached.

According to a third aspect of the present invention, there is provided a medical cart anti-collision detection system, comprising the medical cart anti-collision detection device according to the second aspect of the present invention and a driving record management platform, wherein the medical cart anti-collision detection device is connected to the driving record management platform through a TCP/IP protocol, and when the medical cart enters a coverage area of a wireless local area network, data locally stored in the detection device can be uploaded to the driving record management platform through a touch screen operation of an information display module of the medical cart; the driving record management platform also provides the functions of user management, personal center, alarm recording, health promotion and education and problem consultation.

Has the advantages that:

1. according to the invention, an OBB directed bounding box target detection method is adopted to generate a bounding box of the barrier, and the distance judgment is carried out on the bounding box of the medical trolley, so that the three-dimensional scene is reduced into a two-dimensional scene for modeling, the algorithm complexity is simplified, and the modeling efficiency is improved.

2. The OBB directed bounding box target detection algorithm is matched with the vibration sensor to carry out double detection, so that the problem of accuracy reduction caused by scene dimension reduction is solved, the misjudgment rate is reduced, and the reliability is improved.

3. The driving record management platform can be used for recording original video information acquired by the medical cart in real time and analyzing, processing and recording data. The medical cart driving information, the early warning information and the warning information can be managed in a unified mode conveniently by setting the function of a manager.

4. Data generated in the moving process of the medical cart can be stored in two storage modes (SD card local storage/driving record management platform remote storage). The data storage is safe and reliable, and the checking and unified management are convenient. The medical trolley can achieve early warning, collision recording and the like, and realizes safe advancing and unified management of the medical trolley.

Drawings

Fig. 1 is a flow chart of a medical cart anti-collision detection method of the present invention;

FIG. 2 is a schematic diagram of an OBB detection model used in the present invention;

FIG. 3 is a hardware block diagram of the present invention;

FIG. 4 is a logic diagram of information transmission and storage according to the present invention.

Detailed Description

The technical scheme of the invention is further explained by combining the attached drawings.

According to the method, a driving record management platform of the medical cart is developed through information technologies such as Internet of things, cloud computing and machine learning, early warning, collision record and the like are achieved by utilizing an OBB directed bounding box collision detection technology, and safe advancing and unified management of the medical cart is achieved.

As shown in fig. 1, in one embodiment, the medical cart anti-collision detection method provided by the present invention includes the following steps:

and step S1, acquiring barrier information around the medical cart, and reducing the three-dimensional barrier coordinate information into low-dimensional coordinate information according to a PCA algorithm.

Video acquisition modules are distributed around the medical cart, and the video acquisition modules acquire original video data of the barrier by using the cooperation of a 0V7670 image sensor and a radar ranging sensor and convert the original video data into three-dimensional coordinate information of the barrier after processing. After the three-dimensional coordinate information of the obstacles around the medical cart is acquired, a Principal Component Analysis (PCA) method is adopted, and the high-dimensional data is converted into low-dimensional data by using the idea of reducing the dimensions, so that the data processing complexity is reduced.

The principal component analysis method is a statistical method for reducing dimension, which converts the original random vector related to its component into a new random vector unrelated to its component by means of orthogonal transformation, which is represented algebraically by transforming the covariance matrix of the original random vector into a diagonal matrix and geometrically by transforming the original coordinate system into a new orthogonal coordinate system pointing to p orthogonal directions where sample points are most spread, and then performs dimension reduction processing on a multi-dimensional variable system to convert the multi-dimensional variable system into a low-dimensional variable system with a higher precision.

The calculation steps of the principal component analysis method are as follows:

1. normalization of raw index data: collecting three-dimensional random vector X ═ X1,X2,X3)TN samples xi=(Xi1,Xi2,Xi3)TI is 1,2, …, n, n is more than 3, a sample array is constructed, and the sample array elements are subjected to the following normalized transformation:

Figure BDA0002240731220000031

wherein

Figure BDA0002240731220000032

And obtaining a standardized array W.

2. And (3) solving a correlation coefficient matrix for the normalized matrix W:

Figure BDA0002240731220000033

wherein

Figure BDA0002240731220000034

3. Solving the eigen equation | R- λ I of the sample correlation matrix R3And (3) obtaining characteristic roots, and determining the principal components.

Push button

Figure BDA0002240731220000041

Determining m value to make the utilization rate of information above t (threshold value obtained by deep learning in advance), and determining each characteristic root lambdajJ is 1,2, …, and the equation set Rb is λjb obtaining unit feature vector

Figure BDA0002240731220000042

4. Converting the normalized index variable into principal components:

Figure BDA0002240731220000043

u1 is called the first principal component, U2 is called the second principal component, …, and Up is called the pth principal component.

5. And (3) comprehensively evaluating the m main components:

and carrying out weighted summation on the m principal components to obtain a final evaluation value, wherein the weight is the variance contribution rate of each principal component. The weighted summation is to reduce the dimensionality of the data set while maintaining the features of the data set that contribute most to the variance. The final evaluation value is the projection matrix U*=(U1,…,Um)。

The method can be expressed as:

inputting:

sample set X ═ X1,X2,X3)T(ii) a Low dimensional space dimension m (1 or 2)

The process is as follows:

carrying out standardized transformation on the sample;

computing a covariance matrix W of the samplesTW, and solving a correlation coefficient matrix;

for covariance matrix WTPerforming characteristic value decomposition on W;

obtaining the eigenvector U corresponding to the maximum m eigenvalues1,…,Um

And (3) outputting:

projection matrix U*=(U1,…,Um)

In the present invention, the value of m is generally 2. Projection matrix U*Can be regarded as two-dimensional coordinates after dimension reduction.

In step S2, a bounding box of an obstacle is generated by the OBB directed bounding box object detection method.

OBB (organized Bounding Box, directed Bounding Box): also called orientation bounding box, it will move, zoom, rotate with the object. It is essentially a rectangle closest to the object, except that the rectangle can be rotated arbitrarily according to the first moment of the object. The OBB is closer to the object shape than other collision detection algorithms (such as bounding sphere and AABB), and can significantly reduce the number of bounding volumes. When there are a large number of objects that require collision/intersection detection, this is usually done in two rounds: a first round of rapid testing was done with the surrounding ball and a second round was done with the OBB. Most invisible or unnecessary objects to be cut can be removed in the first round of test, so that the probability of performing the second round of test is greatly reduced, and the detection efficiency is improved. Meanwhile, the result of the OBB bounding box test is more accurate, and the final drawn objects are fewer. Such hybrid enclosures are commonly used in the fields of collision detection, physics/mechanics, and the like. Features of the OBB bounding box: the bounding box area can be minimized under the premise of containing all given points; the generated bounding cartridges have directions.

The OBB directed bounding box collision detection algorithm steps are as follows:

①, randomly selecting n two-dimensional coordinate points { (x1, y1), (x2, y2), …, (xn, yn) } as known points on the boundary of the detection object, wherein the detection object is an obstacle, and the three-dimensional coordinate information of the obstacle can be represented in a two-dimensional coordinate system after dimension reduction in the step S1.

② combine data by coordinate type, x ═ { x1, x2, …, xn }, y ═ y1, y2, …, yn }.

③, respectively calculating the average values averX and averY of the set x and the set y, and constructing a covariance matrix:

Figure BDA0002240731220000051

④, solving the eigenvalue and eigenvector according to the covariance matrix and Jacobian algorithm, selecting the largest eigenvalue as the direction of the OBB bounding box, and solving the eigenvalue and eigenvector according to the covariance matrix by respectively using np.

⑤ a new coordinate system is established according to the obtained characteristic vector, the original data is put back under the new coordinate system, the length, width and central point of the OBB bounding box are obtained, the length, width, central point and the direction of the OBB bounding box form the bounding box of the obstacle, as shown in figure 2, the new coordinate system (X, O, Y) is the coordinate system with the X-axis not horizontal, the bounding box with the X-axis horizontal is the bounding box of the medical cart, B in figure 2 is the bounding box of the medical cart, and A is the bounding box of the obstacle.

And step S3, calculating the distance between the boundary frame of the obstacle and the boundary frame of the medical cart, and performing collision early warning and alarm according to the distance.

Establishing a model according to the obtained coordinate axis of the bounding box of the OBB detection object and the coordinate axis of the bounding box of the medical cart, and judging whether collision occurs: if the projections of the two bounding boxes on all the axes are overlapped, judging that the two bounding boxes collide; otherwise, it is determined that no collision has occurred. The bounding box coordinate axis of the medical cart is a default reference coordinate axis (the x axis is unchanged in the horizontal direction), and the bounding box obtained by the OBB algorithm is only the frame of the obstacle. The model built by two axes is the model shown in fig. 2.

OBB collision detection principle in two-dimensional scene: and respectively projecting the object A and the object B in the X-axis direction and the Y-axis direction, wherein the maximum point coordinate in the Y-axis direction of the object A is Y1, the minimum point coordinate Y2, the minimum point coordinate X2 in the X-axis direction, the maximum point coordinate X4, the maximum point coordinate in the Y-axis direction of the object B is Y3, the minimum point coordinate Y4, the minimum point coordinate X1 in the X-axis direction, and the maximum point coordinate X3. OBB collision detection in a two-dimensional scene has the following rules: the object A and the object B are respectively projected along two coordinate axes, and only under the condition that the two coordinate axes are overlapped, the two objects mean that collision occurs. The line segment X2X3 on the X axis in fig. 2 is the projection overlap of object a and object B on the X axis. The Y-axis has no overlap. Therefore, the logic of the OBB collision detection algorithm is as follows:

(1) the minimum value of the object A in the Y-axis direction is larger than the maximum value of the object B in the Y-axis direction;

(2) the minimum value of the object A in the X-axis direction is larger than the maximum value of the object B in the X-axis direction;

(3) the minimum value of the object B in the Y-axis direction is larger than the maximum value of the object A in the Y-axis direction;

(4) the minimum value of the object B in the X-axis direction is larger than the maximum value of the object A in the X-axis direction;

if the above 4 conditions are satisfied simultaneously, it is verified that the object a and the object B do not coincide, whereas if any one of the above four conditions is not satisfied, it is verified that the object a and the object B coincide. That is, if (Y1-Y4) (Y3-Y2) >0 in the Y-axis direction and (X4-X1) (X3-X2) >0 in the X-axis direction, it is proved that the object A and the object B coincide, otherwise it is proved that the objects A and B do not coincide.

The invention carries out collision risk prompt according to the OBB collision detection principle: when the minimum linear distance between the medical cart boundary frame and the barrier boundary frame is smaller than a certain threshold value, such as 0.2 m, a dangerous condition is judged, a potential collision danger exists, and at the moment, an early warning prompt is triggered, such as an LED lamp flickering early warning.

The method reduces the dimension of the three-dimensional scene into the two-dimensional scene modeling, simplifies the algorithm complexity, but reduces the accuracy to some extent. In order to avoid misjudgment, the vibration sensor can be used for double detection, when the boundary frame of the medical trolley is overlapped with the boundary frame of the surrounding vehicle and the vibration sensor detects vibration, the collision is judged to occur, and the buzzer is triggered to give out a sound alarm.

In this particular scenario of a hospital, an electronic vibration sensor may be used. And the information acquired by the sensor is transmitted to the cloud platform through a special processing chip for data analysis, and is compared with a threshold value obtained by deep learning training to judge whether collision occurs or not, and a signal warning prompt is sent to a processing result. The threshold value refers to the utilization rate t in the step 3 of the principal component analysis method. Different obstacles and different values of the movement states of the obstacles are set so as to obtain the optimal obstacle frame identification efficiency.

According to another embodiment of the invention, a medical cart anti-collision detection system is provided, which is composed of a hardware device and a driving record management platform. Referring to fig. 3, the hardware device portion includes a video capture module, a vibration sensor, an information display module, an information storage module, an alarm module, and the like. The video acquisition modules are distributed around the medical trolley, and the invention adopts a 0V7670 image sensor to cooperate with a radar ranging sensor module to acquire the three-dimensional coordinate information of the barrier. In the prior art, there are many methods for obtaining three-dimensional coordinates of an object by using a camera and radar ranging, and the present invention is not described in detail. The video acquisition module is connected with the STM32 single chip microcomputer through a USB (universal serial bus) to provide original obstacle coordinate data for the anti-collision system of the medical trolley. Vibration sensor passes through the cluster mouth line and is connected with STM32 singlechip, turns into the vibration of sensing into the signal transmission and sends the singlechip to, improves the degree of accuracy that the collision was judged. The information display module is connected with the STM32 single chip microcomputer through a serial port line, can display information such as videos, historical alarms, driving records and the like acquired by the 0V7670 image sensor, and provides an interactive interface for medical cart users. The information storage module is used for storing data locally, when the medical cart enters a coverage range of a wireless local area network, touch screen operation can be performed in the information display module, the locally stored data are uploaded to the driving record management platform, and later maintenance and management are facilitated. The alarm module is connected with the STM32 single chip microcomputer through a serial port line, and is triggered when the alarm condition is met, so that the warning effect is achieved.

The video acquisition module adopts a 0V7670 image sensor, has small volume and low working voltage, provides all functions of a single-chip VGA camera and an image processor, and transmits acquired information to an STM32 single-chip microcomputer through a USB. The method has the advantages of high sensitivity, automatic exposure control, automatic gain control, automatic white balance, automatic elimination of lamplight stripes and the like. The 0V7670 image sensor is matched with the radar ranging sensor module to acquire three-dimensional coordinate information of the obstacle. The collected image information with 30 ten thousand pixels can be directly displayed on a TFT-LCD thin film transistor liquid crystal display. And when the early warning and warning information is triggered, the image information acquired in real time is stored in the SD card and is uploaded to the driving record management platform.

The vibration sensor converts vibration parameters generated during collision into electric signals, and the electric signals are amplified by an electronic circuit and then displayed and recorded. The key point is that the mechanical vibration quantity generated in collision is converted into electric quantity (electromotive force, electric charge and other electric quantities), and then the electric quantity is measured, so as to judge whether collision occurs or not.

The information display module adopts a TFT-LCD (thin film transistor liquid crystal display), and each pixel of the liquid crystal display screen is provided with a Thin Film Transistor (TFT), so that the crosstalk in the non-gating process can be effectively overcome, the static characteristic of the liquid crystal display screen is unrelated to the number of scanning lines, and the image quality is greatly improved.

An information storage module: the data collected by the 0V7670 image sensor and the vibration sensor can be stored in two storage modes, namely an SD card local storage mode and a driving record management platform remote storage mode.

An alarm module: the LED warning device has two functions of LED warning and buzzer warning. When the danger early warning information is triggered, the LED lamp flickers to give an early warning, and when the collision warning information is triggered, the buzzer sounds to give an alarm.

Vehicle driving record management platform: and recording the original video information acquired by the medical cart in real time, and analyzing, processing and recording data. The management function is set, so that the background can conveniently and uniformly manage the driving information, the early warning information and the warning information of the medical cart.

STM32 singlechip: the system is used for calculating and processing images acquired by the 0V7670 image sensor, obtaining coordinate data of the obstacle, judging collision by combining vibration signals transmitted by the vibration sensor, and controlling the display and alarm work of the information display module and the alarm module. The invention selects STM32F107VCT6 with high performance and low power consumption and specially designed RMcortex-M3 kernel for embedded application as the central processing unit of the medical cart collision detection system.

Fig. 4 shows the logic diagram of information transmission and storage of the invention, wherein the calculation refers to an STM32 single chip microcomputer of the hardware structure part of the medical cart, and is responsible for processing the acquired information. The network refers to the Internet using a TCP/IP protocol, and when the medical trolley is in a wireless network coverage environment, collision information, acquired video information, medical trolley use records and the like are uploaded to the driving record management platform through the Internet, so that later maintenance and management are facilitated. Storage refers to an information storage module of the hardware structure part of the medical cart.

The driving record management platform is provided with functions of user management, personal center, alarm recording, health promotion and education, problem consultation and the like. The user management function can allocate different login authorities to different roles such as doctors, nurses, managers, patients, visitors and the like, and standardized operation is facilitated. The personal central function can check personal information, bind personal information, change passwords, check login records, check operation authority, check operation prompts and the like. The warning recording function can check the warning and warning records in days, months, seasons and years, and video data acquired by the medical trolley is stored in the cloud and checked. The health propaganda and education function provides proper propaganda and education information for the user, and the user can acquire practical knowledge on the management platform after the user leaves idle. The question consultation function can provide personalized solutions for user to answer questions and ask puzzles.

The system can perform early warning and recording on potential collision, can also perform warning and recording on collision scraping which occurs, and achieves high-precision and real-time collision detection, early warning and warning. When the boundary frame of the medical trolley and the boundary frame of the barrier are smaller than a certain threshold value, a dangerous condition (potential collision danger exists) is judged, LED lamp flicker early warning in the warning module is triggered, video information acquired by the 0V7670 image sensor is stored in the SD card and uploaded to the driving record management platform, and the driving record management platform records the early warning information. When the medical trolley boundary frame is overlapped with the obstacle boundary frame, the vibration sensor receives vibration information at the same time, and the collision is judged. After collision occurs, a buzzer in the alarm module is triggered to give out sound alarm, video information collected by the 0V7670 image sensor is stored in the SD card and uploaded to the driving record management platform, and the driving record management platform records the alarm information.

11页详细技术资料下载
上一篇:一种医用注射器针头装配设备
下一篇:云图图像预测准确性的评估方法、装置、设备及存储介质

网友询问留言

已有0条留言

还没有人留言评论。精彩留言会获得点赞!

精彩留言,会给你点赞!

技术分类