CN120576619A - An intelligent multi-modal target recognition and fusion sighting device - Google Patents

Abstract

The present invention relates to the field of computer vision, and specifically to an intelligent multimodal target recognition and fusion sighting device connected to a display and control terminal. The device integrates various modules into a housing via a circuit board. The modules include a white light television module, an infrared thermal imager module, a laser ranging module, a power control module, and an intelligent processing component. The power control module is used for power supply, while the white light television module, the infrared thermal imager module, and the laser ranging module perform video acquisition and distance detection. The communication is commanded by a host computer, and the intelligent processing component controls the algorithm flow and processes the acquired data. Finally, the processed video data is displayed via the display and control terminal. The intelligent processing component design of the present invention integrates advanced image enhancement technology, deep learning models and algorithms, and multi-source information fusion strategies to solve the problem of target detection and tracking at night and in low-light environments, so as to cope with the increasingly complex battlefield environment and improve the recognition efficiency and accuracy of images in complex situations.

Description

Intelligent multi-mode target recognition fusion sighting device

Technical Field

The invention relates to the field of computer vision, in particular to an intelligent multi-mode target recognition fusion sighting device.

Background

With the development of modern warfare, targets of enemies, whether guided weapons or defense systems, must be accurately detected and stably tracked over as long a distance as possible to ensure the safety of the present fighter, weaponry and ground facilities. More and more intelligent auxiliary devices are beginning to be assembled into the army, and data from different sensors can be integrated in a complex environment, and clearer and more detailed battlefield images can be provided through target detection and identification of real-time pictures. This helps commander and fighter to better understand the battlefield situation and make accurate decisions.

The application of the existing intelligent auxiliary equipment in the military field greatly improves the battlefield situation awareness capability, particularly the target detection and tracking is significantly advanced, but a plurality of challenges are still faced when the battlefield task is processed in night or low light environment. Namely, when the light condition is bad and the traditional sensor is difficult to capture a clear image, the accurate detection and stable tracking of the target become a great difficulty. In addition, camouflage, interference and high speed movement of objects in complex battlefield environments further exacerbate the complexity of this problem. These factors not only affect the efficiency of the operation, but can also pose a potential threat to the operational safety, as erroneous decisions or delayed responses can result in a reversal of war situation

Disclosure of Invention

In order to solve the technical problems in the prior art, the invention provides an intelligent multi-mode target recognition fusion viewing device, which has the following specific technical scheme:

The intelligent multimode target recognition fusion sighting device is connected with a display control terminal, the device is integrated with each module in a shell through a circuit board, each module comprises a white light television module, a thermal infrared imager module, a laser ranging module, a power control module and an intelligent processing module, the power control module is used for supplying power, the white light television module and the thermal infrared imager module are used for acquiring video signals and transmitting the video signals to the intelligent processing module, the laser ranging module is used for acquiring target distances, the intelligent processing module is connected with the display control terminal through signals, receives external control instructions, performs image preprocessing, target recognition and moving target detection on the video signals according to instruction requirements, and sends recognition detection results to the display control terminal for display.

The intelligent processing assembly comprises an image processing module, a control unit module and an input/output module, wherein the image processing module receives video signals, performs corresponding processing by adopting an image preprocessing, deep learning improved ESRGAN model and a neural network lightweight YoloV algorithm to realize enhancement processing of images and identification and detection of targets in the images, the control unit module receives control instructions, performs resolving analysis and sends Guan Zhiling to a thermal infrared imager module, a white light television module, the image processing module, a laser ranging module and a display control terminal to perform corresponding operation, and the input/output module receives video signals acquired by the thermal infrared imager and the white light television, transmits the processed video signals, and sends processed video signals to the display control terminal, and transmits instruction signals and feedback information.

Furthermore, the laser ranging module sends measured data to the control unit module through the RS422 serial port, the control unit module calculates the received data information, the calculated result is sent to the image processing module, and the image processing module transmits the calculated result to the display control terminal for real-time display.

Further, the input/output module comprises a signal connection socket, a video connection socket and a video conversion board, wherein the signal connection socket and the video connection socket are connected to the display control terminal, the image processing module processes the video signals and then sends AVT video signals to the video conversion board, and the video conversion board converts the AVT signals into SDI video signals and sends the SDI video signals to the display control terminal for display.

Further, the power supply control module comprises a power supply connection socket, a power supply board, a filter and a voltage conversion board, the filter receives external 28V voltage through the power supply connection socket, the 28V voltage is converted into 12V voltage through the power supply board and is respectively supplied to the intelligent processing assembly and the voltage conversion board, the voltage conversion board converts the 12V voltage into 5V voltage to supply power to the laser ranging module, and the intelligent processing assembly supplies 5V voltage to the white light television module and the thermal infrared imager module and supplies 12V power to the video conversion board.

Further, the white light television module mainly comprises two paths of white light televisions consisting of a white light lens and a white light movement, and the thermal infrared imager module mainly comprises an infrared lens, a lens driving plate and an infrared movement.

Further, the white light television module and the thermal infrared imager module transmit video signals to the intelligent processing assembly through the MIPI interface.

Further, the display control terminal sends a signal instruction to the intelligent processing assembly through the CAN/Ethernet.

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

the image processing module adopts a deep learning improved ESRGAN model and a neural network lightweight YoloV algorithm, improves the image effect of the video and the accuracy and precision of target identification, and improves the signal-to-noise ratio, contrast, image resolution and detail texture of the image. The super-resolution and image enhancement tasks of the image are realized, so that the pixel area and texture detail of the weak and small target meet the requirements for processing by a target recognition module.

The image processing module performs optimization and light weight processing through the adopted network model and algorithm, and achieves real-time processing of target identification tasks on the domestic embedded platform.

The sighting device is small in size, easy to carry, suitable for various detection environments, capable of performing function detection immediately after the equipment is subjected to high-low temperature and other tests, and capable of guaranteeing the detection accuracy.

The device is internally provided with a network interface, software and an algorithm can be changed or added through connection, the network and the algorithm are updated, and the product performance is improved in time.

Drawings

Fig. 1 is a software architecture design diagram of an intelligent multi-mode target recognition fusion viewing device according to the present embodiment;

FIG. 2 is a block diagram of the components of the intelligent processing of the present embodiment;

FIG. 3 is a schematic diagram of the power supply of each module of the present embodiment;

fig. 4 is a signal control schematic diagram of each module of the present embodiment.

Detailed Description

In order to make the objects, technical solutions and technical effects of the present invention more apparent, the present invention will be further described in detail with reference to the drawings and examples of the specification.

The intelligent multimode target recognition fusion sighting device comprises a shell, wherein an optical bench is fixedly installed in the shell, and a white light television module, a thermal infrared imager module, a laser ranging module, a power supply control module and an intelligent processing assembly are fixed in the optical bench. The white light television module and the thermal infrared imager module are used for collecting video signals and transmitting the video signals to the intelligent processing assembly. As shown in FIG. 1, the device utilizes a white light television module, a thermal infrared imager module and a laser ranging module to perform video acquisition and distance detection, then uses an upper computer to communicate instructions, performs algorithm flow control and data acquisition processing through an intelligent processing component, and finally displays the processed video data through a display control terminal.

As shown in fig. 2, the intelligent processing assembly includes an image processing module, a control unit module, and an input/output module.

The image processing module receives the video signal and carries out corresponding processing by adopting an image preprocessing and deep learning improved ESRGAN model and a neural network lightweight YoloV algorithm according to the instruction sent by the control unit module, so that the enhancement processing of the image and the identification and detection of the target in the image are realized, and the identification and detection result is sent to the display control terminal to be displayed.

The image processing module obviously reduces the parameter quantity of the model while hardly affecting the performance of the model, improves the characteristic extraction capacity, reduces the convolution structure parameters, improves the fitting capacity of the network, and performs iterative channel pruning on the detection network to realize model compression and acceleration.

The laser ranging module sends the measured data to the control unit module through the RS422 serial port. The control unit module calculates the received data information, sends the calculated result to the image processing module, and the image processing module transmits the calculated result to the display control terminal for real-time display.

The control unit module is divided into an external control and an internal control. In the external control process, an operator CAN send an instruction to the sighting device through CAN communication, the control unit module performs resolving analysis, and sends the phase Guan Zhiling to the thermal infrared imager module, the white light television module, the image processing module and the laser ranging module to perform corresponding operation. In the internal control process, the control on the display function of the received image is mainly performed, and BIT recording, photographing, cross division, picture-in-picture, target positioning and the like are realized.

The input/output module is used for receiving video signals collected by the thermal infrared imager and the white light television, transmitting the processed video signals and sending the processed video signals to the display control terminal. And receiving an external control instruction, and transmitting the control instruction and feedback information.

The input/output module comprises a signal connection socket, a video connection socket and a video conversion board. For example, the white light television module mainly comprises two paths of white light televisions consisting of a white light lens and a white light movement, and the thermal infrared imager module mainly comprises an infrared lens, a lens driving plate and an infrared movement. As shown in fig. 4, when the sighting device works normally, the thermal infrared imager and the two paths of white televisions collect real-time video signals through the machine core, the real-time video signals are transmitted to the intelligent processing assembly through the MIPI interface, the intelligent processing assembly processes the video signals through the three internal functional modules and then sends the AVT video signals to the video conversion board, and the video conversion board converts the AVT signals into SDI video signals and sends the SDI video signals to the display control terminal for display. The display control terminal CAN send a signal instruction to the intelligent processing component through the CAN/Ethernet, and after the intelligent processing component receives the control instruction, the intelligent processing component processes the video signal through the image processing module according to the instruction requirement, processes the image instruction through the control unit module, or sends instruction information to the connecting equipment. The thermal infrared imager, the white light television and the laser range finder respond according to the received instructions.

The power control module is used for providing the voltage required by each module. As shown in fig. 3, the power control module includes a power connection socket, a power board, a filter and a voltage conversion board, and after the sighting device is started, the filter receives an external 28V voltage through the power connection socket and supplies power to each module of the product through the power board. The power panel converts 28V voltage into 12V and respectively supplies the voltage to the intelligent processing assembly and the voltage conversion board, the voltage conversion board converts the 12V voltage into 5V and supplies power to the laser range finder, and the intelligent processing assembly supplies 5V voltage to the white light television module and the thermal infrared imager module and supplies power to the video conversion board by 12V.

The specific use method of the device of the embodiment is as follows:

connecting the signal connection socket and the video connection socket to a display control terminal, and connecting the power connection socket to external power supply equipment;

After video acquisition, displaying the white light video signal on a display control terminal;

the upper computer sends a laser ranging instruction, and the distance data is displayed in a picture of the display control terminal;

The upper computer sends a video display instruction, so that picture-in-picture, single white light or infrared picture images, video font color switching, cross division and the like can be realized;

the upper computer sends an electronic zoom instruction, and the images are displayed in a 2-time and 4-time mode;

the upper computer sends a focusing instruction, and the infrared image can realize automatic or manual zooming;

the upper computer sends a target identification instruction, detects a picture target and marks the target type.

In conclusion, the device integrates technologies such as image sensing, accurate target recognition, laser ranging, infrared thermal imaging and the like, can be integrated into one shell through circuit board design, adopts light-weight portable design, and is specially designed for detecting the demand in situ in real time. The advanced image enhancement technology, the deep learning model and algorithm and the multisource information fusion strategy are designed and fused, the problems of target detection and tracking at night and in low-light environments are solved, the increasingly complex battlefield environments are dealt with, and the recognition efficiency and accuracy of images under the complex conditions are improved. In addition, the device also has strong communication control capability, and can communicate with a thermal infrared imager, a white light television, display control, laser ranging and other modules, so that the optimization of the overall performance and the convenience of operation are ensured.

The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the foregoing detailed description of the invention has been provided, it will be apparent to those skilled in the art that modifications may be made to the embodiments described in the foregoing examples, and that certain features may be substituted for those illustrated and described herein. Modifications, equivalents, and alternatives falling within the spirit and principles of the invention are intended to be included within the scope of the invention.

Claims (8)

1. An intelligent multimodal target recognition and fusion sighting device connected to a display and control terminal, characterized in that the device integrates various modules into a housing via a circuit board. The modules include a white light television module, an infrared thermal imager module, a laser ranging module, a power control module, and an intelligent processing component. The power control module is used to supply power, the white light television module and the infrared thermal imager module are used to acquire video signals and transmit them to the intelligent processing component, the laser ranging module is used to obtain the target distance, and the intelligent processing component is connected to the display and control terminal signal, receives external control commands, performs image preprocessing, target recognition, and moving target detection on the video signal according to the command requirements, and sends the recognition and detection results to the display and control terminal for display. 2. The intelligent multimodal target recognition and fusion sighting device according to claim 1 is characterized in that the intelligent processing component includes an image processing module, a control unit module and an input/output module; the image processing module receives a video signal and uses image preprocessing, a deep learning improved ESRGAN model and a neural network lightweight YoloV5 algorithm to perform corresponding processing to achieve image enhancement processing and recognition and detection of targets in the image; the control unit module receives control instructions, performs solution analysis, and sends relevant instructions to the infrared thermal imager module, the white light television module, the image processing module, the laser ranging module, and the display and control terminal to perform corresponding operations; the input/output module receives video signals collected by the infrared thermal imager and the white light television, transmits the processed video signals, sends the processed video signals to the display and control terminal, and transmits instruction signals and feedback information. 3. The intelligent multimodal target recognition and fusion aiming device according to claim 2 is characterized in that the laser ranging module sends the measured data to the control unit module through the RS422 serial port, the control unit module solves the received data information and sends the solution result to the image processing module, and the image processing module transmits the solution result to the display and control terminal for real-time display. 4. The intelligent multimodal target recognition and fusion sighting device according to claim 2 is characterized in that the input/output module includes a signal connection socket, a video connection socket and a video conversion board, the signal connection socket and the video connection socket are connected to the display and control terminal, the image processing module processes the video signal, and then sends the AVT video signal to the video conversion board, the video conversion board converts the AVT signal into an SDI video signal and sends it to the display and control terminal for display. 5. The intelligent multimodal target recognition and fusion sighting device according to claim 4 is characterized in that the power control module includes a power connection socket, a power board, a filter and a voltage conversion board. After the filter receives an external 28V voltage through the power connection socket, the power board converts the 28V voltage into 12V and supplies it to the intelligent processing component and the voltage conversion board respectively. The voltage conversion board converts the 12V voltage into 5V to power the laser ranging module. The intelligent processing component provides 5V voltage to the white light TV module and the infrared thermal imager module, and supplies 12V to the video conversion board. 6. The intelligent multimodal target recognition and fusion sighting device according to claim 1 is characterized in that the white light television module mainly comprises a two-channel white light television consisting of a white light lens and a white light core, and the infrared thermal imager module mainly comprises an infrared lens, a lens driver board, and an infrared core. 7. The intelligent multimodal target recognition and fusion sighting device according to claim 1, wherein the white light television module and the infrared thermal imager module transmit video signals to the intelligent processing component via a MIPI interface. 8. The intelligent multimodal target recognition and fusion sighting device according to claim 1 is characterized in that the display and control terminal sends signal instructions to the intelligent processing component via CAN/Ethernet.