阅读说明：本技术 一种反伪装智能光电跟踪瞄准镜 (Anti-camouflage intelligent photoelectric tracking sighting telescope ) 是由 李继泉 龙炎 张磊 胡春松 姜立伟 黄泽菁 胡婷 张凯荣 周熙林 张雪娇 于 2021-08-25 设计创作，主要内容包括：本发明公开了一种反伪装智能光电跟踪瞄准镜,包括物镜、镜体等,所述镜体内部安装有偏振电视、激光测距机模块、主控板、电池组件、气压温度传感器和OLED组件,所述主控板包括接口模块、存储模块和核心处理模块,偏振电视输出端、激光测距机模块输出端、OLED组件输入端和外触发线缆信号端均通过接口模块连接在主控板上；存储模块用于储存识别、跟踪算法、分划信息、射表信息,系统运算与控制单元集成在核心处理模块上。本发明瞄准镜采用偏振电视得到目标的多个角度的偏振信息,结合后端的图像处理算法,实现了小型化、轻量化、高效化、低功耗的目标,有较强的应用价值。(The invention discloses an anti-camouflage intelligent photoelectric tracking sighting telescope, which comprises an objective lens, a telescope body and the like, wherein a polarized television, a laser range finder module, a main control board, a battery assembly, an air pressure temperature sensor and an OLED assembly are arranged in the telescope body, the main control board comprises an interface module, a storage module and a core processing module, and the output end of the polarized television, the output end of the laser range finder module, the input end of the OLED assembly and the signal end of an external trigger cable are all connected onto the main control board through the interface module; the memory module is used for storing identification, tracking algorithm, division information and shooting table information, and the system operation and control unit is integrated on the core processing module. The sighting telescope provided by the invention adopts the polarization television to obtain polarization information of multiple angles of the target, combines a rear-end image processing algorithm, realizes the targets of miniaturization, light weight, high efficiency and low power consumption, and has a strong application value.)

1. An anti-camouflage intelligent photoelectric tracking sighting telescope comprises an objective lens (1), laser protective glass (2), a telescope body (3), a key switch assembly (4), an eyepiece lens group (5) and a rotary operation assembly (8); the device is characterized in that the objective lens (1) and the laser protection glass (2) are positioned at the front end of the lens body (3), the key switch assembly (4) and the rotary operation assembly (8) are positioned at the upper end of the lens body (3), and the eyepiece lens group (5) is positioned at the rear end of the lens body (3), and is characterized in that a polarization television (6), a laser range finder module (7), a main control board (9), a battery assembly (10), an air pressure temperature sensor (11) and an OLED assembly (12) are arranged in the lens body (3);

the main control board (9) comprises an interface module, a storage module and a core processing module, and the output end of the polarized television (6), the output end of the laser range finder module (7), the input end of the OLED assembly (12) and the signal end of the external trigger cable (15) are connected to the main control board through the interface module; the memory module is used for storing identification, tracking algorithm, division information and shooting table information, and the system operation and control unit is integrated on the core processing module.

2. The anti-camouflage intelligent photoelectric tracking sighting telescope of claim 1, wherein the division information comprises curves and numbers, and target distance data is read by simply measuring distance according to the size of a target in a visual field.

3. The anti-camouflage intelligent photoelectric tracking sighting telescope according to claim 1 or 2, wherein the shooter information comprises distance data, temperature data and air pressure data, the system acquires real-time data through the laser range finder module (7) and the air pressure temperature sensor (11), and the shooting information is combined to automatically mount the shooting.

4. The anti-camouflage intelligent photoelectric tracking sighting telescope according to claim 3, wherein the anti-camouflage intelligent photoelectric tracking sighting telescope comprises a simple fire control system, the simple fire control system comprises a laser range finder module (7) and an air pressure temperature sensor (11), the target distance information is acquired, the environmental temperature and air pressure data are combined, the meter mounting position is calculated according to the trajectory information, and the target distance information is overlaid and displayed on a display screen of the OLED assembly (12) through an imaging system.

5. The anti-camouflage intelligent photoelectric tracking sighting telescope of claim 4, wherein the system operation on the core processing module comprises a target recognition and contour extraction algorithm, and the implementation process of the algorithm comprises the following steps:

step 1: the input image of one frame is divided into N × N network units by using a CNN network of YOLO, and then each cell is responsible for detecting targets with central points falling in the cell;

step 2: processing an input image in grids, wherein each grid unit can predict B frames, including the confidence of each frame being a target and the probability of each frame region on a plurality of categories; the confidence level includes two aspects: firstly, the probability size of the boundary box containing the target, and secondly, the accuracy of the boundary box; the former is denoted as pr (object), and when the bounding box is background and has no object, pr (object) is 0; when the bounding box includes a target, pr (object) is 1; the accuracy of the latter bounding box can be characterized by the intersection ratio IOU (intersection over unit) of the predicted box and the actual box, and is marked asConfidence may therefore be defined asThe size and position of the bounding box can be characterized by 4 values: (x, y, w, h), where (x, y) is the center coordinates of the bounding box, and w and h are the width and height of the bounding box; the predicted value (x, y) of the center coordinate is the offset value relative to the upper left corner coordinate point of each cell, while the predicted values of w and h of the bounding box are the ratio of width to height relative to the entire picture, so theoretically the size of the 4 elements should be [0,1 ]]A range; thus the predicted value of each bounding box actually contains 5 elements, namely (x, y, w, h, c), of which the first 4 characterize the size and position of the bounding box, and the last value is the confidence;

each network cell also predicts probability values for each class, which characterize the probability that the bounding box for which the cell is responsible for predicting belongs to each class, and these probabilities are conditional probabilities at each bounding box confidence, namely Pr (classi | object);

step 3: according to Step2, N × B target windows can be predicted, and then the target windows with lower possibility are removed according to the initially set threshold condition;

step 4: based on the operation of Step3, redundant cell removal is carried out by using a non-maximum suppression algorithm, which comprises the following steps: firstly, finding the box with the highest confidence coefficient from all the detection boxes, then calculating the IOU of the rest of the detection boxes one by one, if the IOU value is larger than a set threshold value, which indicates that the coincidence degree is too high, removing the box, and then repeating the operation on the rest of the detection boxes until all the detection boxes are processed;

step 5: and after the redundant unit is removed, performing target contour extraction enhancement on the target image by using a high-precision target contour extraction algorithm.

6. The anti-disguising intelligent photoelectric tracking sighting telescope according to claim 5, wherein the telescope body (3) is fixedly connected with the interface assembly (14) through a bolt, the interface assembly (14) is provided with a locking handle (13), and the external trigger cable (15) is connected to the telescope body (3) through a connector.

7. Disguised intelligent electro-optical tracking telescope according to claim 6, characterized in that an eye mask (16) is mounted on the rear side of the eyepiece group (5).

Technical Field

The invention belongs to the field of sighting telescope, and particularly relates to an intelligent photoelectric tracking sighting telescope with an anti-camouflage function.

Technical Field

The photoelectric sighting telescope on the market at present has two kinds of forms of single channel and multichannel integration, and single channel photoelectric sighting telescope roughly can be divided into white light sighting telescope, digital shimmer sighting telescope, infrared sighting telescope. The white light sighting telescope has sharp imaging and rich details in a good light environment, but has low informatization degree, basically depends on the literacy of an observer for disguising a target, and does not have automatic disguising and identifying capability; digital low-light sighting telescope and infrared sighting telescope can be used in dark environment, but high-frequency optical spectrum information is insufficient, imaging quality details are not abundant, layers are not good enough, camouflage recognition capability is limited, and the difficulty in realizing an intelligent recognition tracking function is large.

The multi-channel integration mainly adopts different imaging systems of white light and low light or white light and infrared to carry out physical combination, light rays with different wave bands respectively enter corresponding objective lenses, an imaging device finishes image acquisition and processing, and finally, observation and aiming are finished through a shared eyepiece. The multi-channel integration mode realizes the fusion and complementation of frequency spectrum information to a certain extent, and in a complex environment, the effectiveness of identification, tracking, aiming and the like of a camouflage target is improved to a certain extent, but the identification capability still further improves the space, and the volume is larger and the weight is heavier.

Disclosure of Invention

In order to solve the problems, the invention provides the intelligent photoelectric tracking sighting telescope which is small in size, light in weight and has the anti-camouflage function, the sighting telescope adopts a polarization television to obtain polarization information of a target at multiple angles, and the sighting telescope is combined with a rear-end image processing algorithm, so that the probability of finding and automatically identifying the target is improved, the capability of camouflage identification is realized, and the anti-camouflage capability of equipment is improved; the problems of large size, heavy weight and the like of the conventional fusion sighting telescope are solved, the goals of miniaturization, light weight, high efficiency and low power consumption are realized, and the fusion sighting telescope has a high application value.

The invention is realized by adopting the following technical scheme:

an anti-camouflage intelligent photoelectric tracking sighting telescope comprises an objective lens, laser protective glass, a telescope body, a key switch assembly, an eyepiece group and a rotary operation assembly; the utility model discloses a laser telescope, including objective, laser protection glass, key switch subassembly, rotatory operation subassembly, eyepiece group, laser range finder module, main control board, battery pack, atmospheric pressure temperature sensor and OLED subassembly, objective, laser protection glass are located the mirror body front end, key switch subassembly, rotatory operation subassembly are located the mirror body upper end, the eyepiece group is located mirror body rear end, its characterized in that, mirror body internally mounted has polarization TV, laser range finder module, main control board, battery pack, atmospheric pressure temperature sensor and OLED subassembly.

The main control board comprises an interface module, a storage module and a core processing module, and the output end of the polarized television, the output end of the laser range finder module, the input end of the OLED assembly and the signal end of the external trigger cable are all connected to the main control board through the interface module; the memory module is used for storing identification, tracking algorithm, division information and shooting table information, and the system operation and control unit is integrated on the core processing module.

Further, the division information includes a curve and a number, and simple ranging can be performed by comparing the size of the target in the field of view, and target distance data can be read.

Furthermore, the shooting table information comprises distance data, temperature data and air pressure data, the system can acquire real-time data through the laser range finder module and the air pressure temperature sensor, and the shooting table information is combined to automatically mount the fixed shooting table.

Furthermore, anti-camouflage intelligent photoelectric tracking sighting telescope includes a simple and easy fire control system, and this simple and easy fire control system includes a laser range finder module and an atmospheric pressure temperature sensor, can acquire target distance information rapidly, combines ambient temperature, atmospheric pressure data, then calculates the dress table position according to trajectory information to through the display screen stack of imaging system display in OLED subassembly.

Further, the system operation on the core processing module comprises a target identification and contour extraction algorithm, and the algorithm implementation process comprises the following steps:

step 1: the input image of one frame is divided into N × N network units by using a CNN network of YOLO, and then each cell is responsible for detecting targets with central points falling in the cell;

step 2: processing an input image in grids, wherein each grid unit can predict B frames, including the confidence of each frame being a target and the probability of each frame region on a plurality of categories; the confidence level includes two aspects: one is the size of the likelihood that the bounding box contains the target, and the accuracy of this bounding box. The former is denoted as pr (object), and when the bounding box is background and has no object, pr (object) is 0; when the bounding box includes an object, pr (object) is 1. The accuracy of the latter bounding box can be characterized by the intersection ratio IOU (intersection over unit) of the predicted box and the actual box, and is marked asConfidence may therefore be defined asThe size and position of the bounding box can be characterized by 4 values: (x, y, w, h), where (x, y) is the center coordinates of the bounding box, and w and h are the width and height of the bounding box. The predicted value (x, y) of the center coordinate is the offset value relative to the upper left corner coordinate point of each cell, while the predicted values of w and h of the bounding box are the ratio of width to height relative to the entire picture, so theoretically the size of the 4 elements should be [0,1 ]]And (3) a range. Thus the predicted value for each bounding box actually contains 5 elements, namely (x, y, w, h, c), of which the first 4 characterize the size and position of the bounding box, and the last value is the confidence.

Each network cell also predicts probability values for each class, which characterize the probability that the bounding box for which the cell is responsible for predicting belongs to each class, and these probabilities are the conditional probabilities at each bounding box confidence, namely Pr (classi | object).

Step 3: according to Step2, N × B target windows can be predicted, and then the target windows with lower possibility are removed according to the threshold condition which is initially set.

Step 4: based on the operation of Step3, redundant cell removal is carried out by using a non-maximum suppression algorithm, which comprises the following steps: the box with the highest confidence is found from all the detection boxes, the IOU of the rest of the detection boxes is calculated one by one, if the IOU value is larger than a set threshold value, which indicates that the coincidence degree is too high, the box is removed, and the operation is repeated for the rest of the detection boxes until all the detection boxes are processed.

Step 5: and after the redundant unit is removed, performing target contour extraction enhancement on the target image by using a high-precision target contour extraction algorithm.

Furthermore, the mirror body is fixedly connected with the interface assembly through a bolt, the interface assembly is provided with a locking handle, and the external trigger cable is connected to the mirror body through a connector.

Furthermore, an eyeshade is arranged on the rear side of the eyepiece group.

Compared with the prior art, the invention has the advantages that:

(1) by adopting a mode of a conventional imaging lens and a polarization television, the polarization television separates light rays with different polarization angle directions according to the polarization characteristics of the light rays and the polarization grating principle, 4-angle polarized light information can be simultaneously obtained by one-time exposure, and the polarization angle and the polarization state image in the current view field can be separated by combining image resolving and fusion processing at the rear end, so that a television signal with polarization information is finally formed. The light intensity image of the target reflected light after being modulated by the polarizing optical device is used for obtaining the polarization characteristic image of the light beam, more important characteristics such as materials, geometric shapes, surface roughness and the like of objects in a scene can be provided, background and other invalid information are filtered, the image quality is enhanced, the human eye sensory effect and the camouflage recognition capability are improved, and 'anti-camouflage' is realized.

(2) The polarized television can detect the range of the wavelength of light wave of 400-1100 nm; the existing common sniper is attached with a photoelectric product, the wavelength of a used laser is mostly 800-1064 nm, and the polarization television adopted in the anti-camouflage intelligent photoelectric tracking sighting telescope can detect and discover the photoelectric product emitting laser spectrum, quickly discover the enemy camouflage sniper and further realize 'anti-camouflage'.

(3) By adopting the polarization television, the sighting telescope has the advantages of reduced volume and weight and high zero stability of the aiming point.

(4) The intelligent recognition and tracking function is realized through the target contour extraction and recognition technology.

Drawings

FIG. 1 is a schematic block diagram of an anti-camouflage intelligent photoelectric tracking sighting telescope;

FIG. 2 is an appearance view of an anti-camouflage intelligent photoelectric tracking sighting telescope;

FIG. 3 is a view showing the internal structure of an anti-camouflage intelligent photoelectric tracking sighting telescope;

FIG. 4 is an architectural diagram of a target contour extraction enhancement and recognition algorithm.

In the figure: 1. objective lens, 2, laser protection glass, 3, the mirror body, 4, key switch subassembly, 5, the eyepiece group, 6, polarization TV, 7, laser range finder module, 8, rotation operation subassembly, 9, main control board, 10, battery pack, 11, atmospheric pressure temperature sensor, 12, OLED subassembly, 13, locking handle, 14, interface module, 15, external trigger cable, 16, eye-shade.

The specific implementation mode is as follows:

the technical solution in the embodiments of the present invention is clearly and completely described below with reference to the accompanying drawings.

As shown in fig. 1-3, an anti-camouflage intelligent photoelectric tracking sighting telescope comprises an objective lens 1, laser protection glass 2, a telescope body 3, a key switch assembly 4, an eyepiece lens group 5, a polarization television 6, a laser range finder module 7, a rotary operation assembly 8, a main control board 9, a battery assembly 10, an air pressure temperature sensor 11, an OLED assembly 12, a locking handle 13, an interface assembly 14, an external trigger cable 15 and an eye shield 16.

As shown in fig. 2 and 3, the objective lens 1 and the laser protection glass 2 are positioned at the front end of the lens body 3, the key switch assembly 4 and the rotary operation assembly 8 are positioned at the upper end of the lens body 3, the eyepiece lens group 5 is positioned at the rear end of the lens body 3, and the rear side of the eyepiece lens group 5 is provided with an eye patch 16; the inside of the mirror body 3 is provided with a polarized television 6, a laser range finder module 7, a main control board 9, a battery assembly 10, an air pressure temperature sensor 11 and an OLED assembly 12; the endoscope body 3 is fixedly connected with an interface component 14 through a bolt, a locking handle 13 is arranged on the interface component 14, and an external trigger cable 15 is connected to the endoscope body 3 through a connector; through the cooperation of two control units of key switch subassembly 4 and rotary switch subassembly 8, can carry out human-computer interaction with the gun sight.

The main control board 9 comprises an interface module, a storage module and a core processing module, and the output end of the polarized television 6, the output end of the laser range finder module 7, the input end of the OLED assembly 12 and the signal end of the external trigger cable 15 are connected to the main control board through the interface module; the memory module is used for storing identification, tracking algorithm, division information and shooting table information, and the system operation and control unit is integrated on the core processing module.

The division information comprises curves and numbers, simple distance measurement can be carried out by comparing the sizes of targets in the view field, and target distance data can be read; the shooting table information comprises distance data, temperature data and air pressure data, the system can acquire real-time data through the laser range finder module 7 and the air pressure temperature sensor 11, and the shooting table information is combined to automatically install the shooting table.

The objective lens 1 is positioned at the front end of the lens body 3 and is imaged on a detector image surface of the polarization television 6; the polarized television 6 modulates and collects the polarization state of light beams through a micro-nano grating structure on the area array detector according to the polarization characteristics of the light, 4-degree polarized light information can be obtained simultaneously through one-time exposure, and the polarization angle and the polarization state image in the current view field are separated by combining image resolving and fusion processing at the rear end to form a polarized television signal. Stray light elimination is realized, the target feature recognition rate is improved, background and other invalid information can be filtered, the image quality is enhanced, the target recognition capability is effectively improved, and the human eye sensory effect and camouflage recognition capability are improved.

The polarized light television can detect light waves with the wavelength of 400-1100 nm, and the working wavelength range of a laser range finder module in the existing sighting telescope is covered by 800-1064 nm. In an actual battlefield, when a camouflaged sniper measures the distance of a live target through a laser range finder module of the sighting telescope, the camouflaged intelligent photoelectric tracking sighting telescope can detect and discover the distance measuring light spot, so that an enemy sniper can be rapidly discovered.

The signal of the polarized television 6 is transmitted to the main control board 9, the image information is transmitted to the OLED assembly 12 after being processed by the algorithm, at the moment, the electric signal is converted into a light signal to be displayed on the OLED12 screen, and finally, the image information is transmitted to human eyes through the eyepiece assembly 5.

As shown in fig. 4, in the architecture of the target contour extraction enhancement and recognition algorithm, the processing algorithm includes a target recognition and contour extraction algorithm, and the algorithm implementation process includes the following steps:

step 1: the input image of one frame is divided into N × N network units by using a CNN network of YOLO, and then each cell is responsible for detecting targets with central points falling in the cell;

step 2: processing an input image in grids, wherein each grid unit can predict B frames, including the confidence of each frame being a target and the probability of each frame region on a plurality of categories; the confidence level includes two aspects: one is the size of the likelihood that the bounding box contains the target, and the accuracy of this bounding box. The former is denoted as pr (object), and when the bounding box is background and has no object, pr (object) is 0; when the bounding box includes an object, pr (object) is 1. The accuracy of the latter bounding box can be characterized by the intersection ratio IOU (intersection over unit) of the predicted box and the actual box, and is marked asConfidence may therefore be defined asThe size and position of the bounding box can be characterized by 4 values: (x, y, w, h), where (x, y) is the center coordinates of the bounding box, and w and h are the width and height of the bounding box. The predicted value (x, y) of the center coordinate is relative toThe offset value of the upper left coordinate point of each cell, and the predicted values of w and h for the bounding box are the ratio of width to height relative to the entire picture, so that theoretically the size of the 4 elements should be [0,1 ]]And (3) a range. Thus the predicted value for each bounding box actually contains 5 elements, namely (x, y, w, h, c), of which the first 4 characterize the size and position of the bounding box, and the last value is the confidence.

Each network cell also predicts probability values for each class, which characterize the probability that the bounding box for which the cell is responsible for predicting belongs to each class, and these probabilities are the conditional probabilities at each bounding box confidence, namely Pr (classi | object).

Step 3: according to Step2, N × B target windows can be predicted, and then the target windows with lower possibility are removed according to the threshold condition which is initially set.

Step 4: based on the operation of Step3, redundant cell removal is carried out by using a non-maximum suppression algorithm, which comprises the following steps: the box with the highest confidence is found from all the detection boxes, the IOU of the rest of the detection boxes is calculated one by one, if the IOU value is larger than a set threshold value, which indicates that the coincidence degree is too high, the box is removed, and the operation is repeated for the rest of the detection boxes until all the detection boxes are processed.

Step 5: and after the redundant unit is removed, performing target contour extraction enhancement on the target image by using a high-precision target contour extraction algorithm.

The anti-camouflage intelligent photoelectric tracking sighting telescope comprises a simple fire control system consisting of a laser range finder module 7 and an air pressure temperature sensor 11, can rapidly acquire target distance information by controlling a trigger switch in an external trigger cable 15, combines environmental temperature and air pressure data, calculates a meter mounting position according to trajectory information, and displays the meter mounting position on a display screen of an OLED assembly 12 in an overlapped mode through an imaging system. The shooter only needs to observe through the eyepiece component 5 and divide the mounted table to aim at the target shooting, so that the calculation workload of the shooter is greatly reduced, and the reaction speed is improved.

The working principle is as follows: after the sighting telescope is installed in a battery 18650, the battery cover is locked, and a power switch key in the key switch assembly 4 is pressed for a long time to start the sighting telescope. Operating the rotary operation component 8, selecting a system working mode, when switching to an intelligent mode, imaging a scene in a field of view on a detector image surface of the polarized television 6 through an optical system of the objective lens 1, then transmitting image information to the main control board 9, processing the image information by the main control board 9 through an identification and tracking algorithm, automatically identifying a camouflaged person and a vehicle target, transmitting the processed image to the OLED component 12 for image display, and observing and identifying human eyes by adhering to the eye mask 16 through the eyepiece component 5; after the hit target is selected, a trigger switch in the external trigger cable 15 is operated to measure the distance of the selected target, the system completes automatic table installation according to the distance information, the temperature information and the air pressure information to generate a shooting division, and the shooting division is aligned to the selected target to shoot. When a knob switch in the rotary operation assembly 8 is operated to switch the mode to a manual mode, division information containing curves and numbers can be called out, simple distance measurement can be carried out through comparing the size of a target in a visual field, target distance data can be read, and manual meter installation can be completed by combining temperature and air pressure information. By further operating the knob switch in the rotational operation member 8, the setting of parameters such as brightness and display mode can be performed.

The whole sighting telescope is connected with the gun through an interface assembly 14 and a locking handle 13, the whole sighting telescope is powered by a 18650 battery, a key switch assembly is turned on and off, and a rotary operation assembly is used for selecting a working mode, setting parameters and the like.