Target tracking identification circuit based on SOC

Abstract

The invention discloses an SOC-based target tracking identification circuit, and relates to the technical field of electronics. The SOC-based target tracking identification circuit comprises a calculation and processing unit, a power supply and control unit, a storage and cache unit and a receiving and sending unit, and the calculation and processing unit comprises a 7045-SOC chip and an Atlas200 acceleration module; and the receiving and transmitting unit comprises a camera link receiving and transmitting connector, an SDI receiving and transmitting connector, a multi-path MT optical fiber interface, an HDMI interface socket and a network interface socket. According to the target tracking identification circuit based on the SOC, the camera link in / out IP module, the SDI3G in / out IP module and the 8b10b IP module are integrated, corresponding hardware chips are omitted, and reusability is enhanced. According to the circuit design, a camera link receiving and transmitting connector, an SDI receiving and transmitting connector, a multi-path MT optical fiber interface, an HDMI interface socket and a network interface socket are integrated, and input of various types of video data and an output mode of selective display can be achieved.

Description

Technical Field

[0001] The invention relates to the technical field of electronic technology, and in particular to a target tracking and identification circuit based on SOC. Background Art

[0002] Currently, target tracking and recognition are widely used in optoelectronic systems. The image data collected by sensors with various interfaces are preprocessed, and then the data is processed by the tracking algorithm after selection, and the azimuth vector is output to the servo to achieve target tracking; the preprocessed image data is passed through the recognition algorithm, and the ID number and the target's four-corner coordinate information are output. Finally, the image data is superimposed with the interface and tracking frame, and the format is converted and output to the terminal display for observation.

[0003] Optoelectronic equipment is gradually developing towards intelligence and lightweight. In traditional solutions, target recognition relies on high-computing computing modules such as graphics cards. It is impossible to place a high-computing computing board alone in the current optoelectronic chassis or cabin. In traditional solutions, various types of video data need to be matched with corresponding interface codec chips, which increases the dependence on hardware and the size of the hardware.

[0004] In summary, the present invention provides a target tracking and recognition circuit based on SOC, which is used to solve the problem that target recognition in traditional solutions relies on high-computing power computing modules such as graphics cards, increasing the dependence on hardware and the hardware size. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention provides a target tracking and recognition circuit based on SOC, which solves the problem that optoelectronic equipment is gradually developing towards intelligence and lightweight. In traditional solutions, target recognition relies on high-computing power computing modules such as graphics cards. It is impossible to place a high-computing power computing board alone inside the current optoelectronic chassis or cabin; in traditional solutions, various types of video data need to be matched with corresponding interface codec chips, which increases the degree of dependence on hardware and the problem of hardware size.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: a target tracking and identification circuit based on SOC, comprising:

[0007] A computing and processing unit, the computing and processing unit includes a 7045-SOC chip and an Atlas200 acceleration module. The 7045-SOC chip is internally provided with a cameralinkin / out IP module, an SDI_3Gin / out IP module and an 8b10bIP module. The 7045-SOC chip cooperates with the Atlas200 acceleration module to realize the functions of data encoding and decoding, target tracking, recognition calculation, image character data superposition, and data encoding display output of the entire system;

[0008] Power supply and control unit to achieve stable and sequential power supply of circuit boards;

[0009] Storage and cache unit, to realize data cache and storage requirements during operation;

[0010] The receiving and sending unit includes a cameralink transceiver connector, an SDI transceiver connector, a multi-channel MT optical fiber interface, an HDMI interface socket, and a network interface socket, which are used to realize the image data receiving and sending requirements during operation.

[0011] Preferably, the cameralink in / out IP module, SDI_3G in / out IP module and 8b10b IP module exchange data with the outside world through the cameralink transceiver connector, the SDI transceiver connector and the multi-channel MT optical fiber interface respectively.

[0012] Preferably, the HDMI interface socket is used to receive and transmit high-definition digital video and audio signals, and the network interface socket is used to achieve connection of network signals.

[0013] Preferably, the multi-channel MT optical fiber interface is divided into two single-mode optical modules and one multi-mode optical module, the cameralink transceiver connector is provided with 4 groups of Cameralink external transceiver interfaces, the SDI transceiver connector is provided with 4 SDI receptions, the HDMI interface socket is provided with 2, and the network interface socket is provided with 2.

[0014] Preferably, the Atlas200 acceleration module is physically connected to the 7045-SOC chip via PCIE.

[0015] Preferably, the Atlas200 acceleration module can deploy the yolov8 neural network algorithm to run target recognition.

[0016] Preferably, the power supply and control unit includes a 4644 power supply module, a 51200 power supply module and a 706 monitoring module. .

[0017] Preferably, the storage and cache unit includes a DDR3 cache chip, an EMMC storage chip, an SD card memory and an external USB storage interface.

[0018] Preferably, the EMMC storage chip and SD card memory are 64G-emmc and 256G-SD card respectively.

[0019] The present invention discloses a target tracking and identification circuit based on SOC, which has the following beneficial effects:

[0020] 1. Compared with the interface chips used in the data encoding and decoding of the traditional cameralink, sdi, and optical fiber acquisition parts, this SOC-based target tracking and recognition circuit is based on the integration of cameralink in / out IP modules, SDI_3G in / out IP modules, and 8b10b IP modules, eliminating the corresponding hardware chips and enhancing reusability. The circuit design integrates cameralink transceiver connectors, SDI transceiver connectors, multi-channel MT optical fiber interfaces, HDMI interface sockets, and network interface sockets, which can realize various types of video data input and select display output modes. The circuit design integrates the Atlas200 acceleration module, which is physically connected to the 7045-SOC chip through the PCIE method, and realizes fast data interaction through the AXI bus based on the XDMA IP, greatly improving the computing power of the board.

[0021] 2. The target tracking and identification circuit based on SOC adopts small package components in circuit design and optimizes the board-level layout. The size of the entire circuit board is 110mm*140mm*2.5mm. The entire circuit adopts 5-14V adaptive power input and includes current monitoring function;

[0022] It includes 4 sets of Cameralink external transceiver interfaces, 4 SDI receiving and 4 SDI sending interfaces, two single-mode optical modules and one multi-mode optical module, 2 HDMI interfaces, and 2 network interfaces. It can adapt to the input of various data formats and select the format of data output. It includes Atlas200 acceleration module, can deploy yolov8 neural network algorithm to run target recognition, and includes 64G-emmc and 256G-SD cards for on-board data storage. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0024] Figure 1 This is a functional schematic diagram of the circuit board of the present invention;

[0025] Figure 2 This is a flow chart of the target tracking and identification scheme implemented by the circuit board of the present invention. DETAILED DESCRIPTION

[0026] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention are clearly and completely described. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0027] The embodiment of the present application solves the problem that optoelectronic equipment is gradually developing towards intelligence and lightweight by providing a target tracking and recognition circuit based on SOC. In the traditional solution, target recognition relies on high-computing power computing modules such as graphics cards, and it is impossible to place a high-computing power computing board alone inside the current optoelectronic chassis or cabin; in the traditional solution, various types of video data need to be matched with corresponding interface codec chips, which increases the degree of dependence on hardware and the problem of hardware size.

[0028] Compared with the interface chips used for data encoding and decoding of traditional cameralink, sdi, and optical fiber acquisition parts, this technology is based on the integration of cameralink in / out IP modules, SDI_3G in / out IP modules, and 8b10b IP modules, eliminating the corresponding hardware chips and enhancing reusability. The circuit design integrates cameralink transceiver connectors, SDI transceiver connectors, multi-channel MT optical fiber interfaces, HDMI interface sockets, and network interface sockets, which can realize various types of video data input and select display output methods. The circuit design integrates Atlas200 acceleration module, which is physically connected to the 7045-SOC chip through PCIE, and realizes fast data interaction through the AXI bus based on XDMA IP, greatly improving the computing power of the board.

[0029] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0030] Embodiment 1:

[0031] According to the instruction manual Figure 1 , 2 As shown, an embodiment of the present invention discloses a target tracking and identification circuit based on SOC.

[0032] It includes a computing and processing unit, which includes a 7045-SOC chip and an Atlas200 acceleration module. The 7045-SOC chip is internally provided with a cameralink in / out IP module, an SDI_3G in / out IP module and an 8b10b IP module. The 7045-SOC chip cooperates with the Atlas200 acceleration module to realize the functions of data encoding and decoding, target tracking, recognition calculation, image character data superposition, and data encoding display output of the entire system;

[0033] Power supply and control unit to achieve stable and sequential power supply of circuit boards;

[0034] Storage and cache unit, to realize data cache and storage requirements during operation;

[0035] The receiving and sending unit includes a cameralink transceiver connector, an SDI transceiver connector, a multi-channel MT optical fiber interface, an HDMI interface socket, and a network interface socket, which are used to realize the image data receiving and sending requirements during operation.

[0036] Compared with the interface chips used for data encoding and decoding of traditional cameralink, sdi, and optical fiber acquisition parts, this technology is based on the integration of cameralink in / out IP modules, SDI_3G in / out IP modules, and 8b10b IP modules, eliminating the corresponding hardware chips and enhancing reusability. The circuit design integrates cameralink transceiver connectors, SDI transceiver connectors, multi-channel MT optical fiber interfaces, HDMI interface sockets, and network interface sockets, which can realize various types of video data input and select display output methods. The circuit design integrates Atlas200 acceleration module, which is physically connected to the 7045-SOC chip through PCIE, and realizes fast data interaction through the AXI bus based on XDMA IP, greatly improving the computing power of the board.

[0037] The target tracking and identification circuit module solution integrates advanced computing and processing, stable and reliable power supply and control, efficient and flexible storage and caching, and diverse receiving and transmitting interfaces. It can realize real-time tracking and identification of targets in images and meet the needs of various application scenarios.

[0038] The 7045-SOC chip and Atlas200 acceleration module decode the received image data for subsequent processing; using advanced algorithms and acceleration modules, the targets in the image are tracked and identified in real time. Relevant character information, such as target name and recognition probability, is superimposed on the identified target. The processed data is encoded into a format suitable for display and output to the display device.

[0039] Specifically disclosed, the cameralink in / out IP module, the SDI_3G in / out IP module and the 8b10bIP module exchange data with the outside world through the cameralink transceiver connector, the SDI transceiver connector and the multi-channel MT optical fiber interface respectively.

[0040] Furthermore, the HDMI interface socket is used to receive and transmit high-definition digital video and audio signals, and the network interface socket is used to achieve the connection of network signals.

[0041] Particularly disclosed, the multi-channel MT optical fiber interface is divided into two single-mode optical modules and one multi-mode optical module, the cameralink transceiver connector is provided with 4 groups of Cameralink external transceiver interfaces, the SDI transceiver connector is provided with 4 SDI receptions, the HDMI interface socket is provided with 2, and the network interface socket is provided with 2.

[0042] Furthermore, the Atlas200 acceleration module is physically connected to the 7045-SOC chip via PCIE.

[0043] Specifically disclosed, the Atlas200 acceleration module is capable of deploying the yolov8 neural network algorithm to run target recognition.

[0044] It is particularly disclosed that the power supply and control unit includes a 4644 power module, a 51200 power module and a 706 monitoring module.

[0045] The 4644 power module, 51200 power module and 706 monitoring module provide stable and reliable power supply to ensure the normal operation of the entire circuit module; supply power to each part in a preset sequence to avoid current shock and damage; and monitor the power supply status in real time through the 706 monitoring module to ensure that timely measures are taken to protect the circuit under abnormal circumstances.

[0046] Embodiment 2:

[0047] Embodiment 1:

[0048] According to the instruction manual Figure 1 , 2 As shown, an embodiment of the present invention discloses a target tracking and identification circuit based on SOC.

[0049] It includes a computing and processing unit, which includes a 7045-SOC chip and an Atlas200 acceleration module. The 7045-SOC chip is internally provided with a cameralink in / out IP module, an SDI_3G in / out IP module and an 8b10b IP module. The 7045-SOC chip cooperates with the Atlas200 acceleration module to realize the functions of data encoding and decoding, target tracking, recognition calculation, image character data superposition, and data encoding display output of the entire system;

[0050] Power supply and control unit to achieve stable and sequential power supply of circuit boards;

[0051] Storage and cache unit, to realize data cache and storage requirements during operation;

[0052] The receiving and sending unit includes a cameralink transceiver connector, an SDI transceiver connector, a multi-channel MT optical fiber interface, an HDMI interface socket, and a network interface socket, which are used to realize the image data receiving and sending requirements during operation.

[0053] Furthermore, the storage and cache unit includes a DDR3 cache chip, an EMMC storage chip, an SD card memory and an external USB storage interface.

[0054] Furthermore, the EMMC storage chip and SD card memory are 64G-emmc and 256G-SD card respectively.

[0055] DDR3 cache chip, EMMC memory chip, SD card storage and external USB storage interface DDR3 cache chip is used to quickly store and read temporary data to improve processing speed. Important data and configuration files are saved through EMMC memory chip and SD card storage to ensure long-term data preservation. USB storage interface is provided to facilitate data backup and transmission for users.

[0056] Compared with the interface chips used for data encoding and decoding of traditional cameralink, sdi, and optical fiber acquisition parts, this technology is based on the integration of cameralink in / out IP modules, SDI_3G in / out IP modules, and 8b10b IP modules, eliminating the corresponding hardware chips and enhancing reusability. The circuit design integrates cameralink transceiver connectors, SDI transceiver connectors, multi-channel MT optical fiber interfaces, HDMI interface sockets, and network interface sockets, which can realize various types of video data input and select display output methods. The circuit design integrates Atlas200 acceleration module, which is physically connected to the 7045-SOC chip through PCIE, and realizes fast data interaction through the AXI bus based on XDMA IP, greatly improving the computing power of the board.

[0057] The circuit design uses small package devices and optimizes the board-level layout. The size of the entire circuit board is 110mm*140mm*2.5mm. The entire circuit uses a 5-14V adaptive power input, including a current monitoring function. It contains 4 sets of Cameralink external transceiver interfaces, 4 SDI receiving and 4 SDI sending interfaces, two single-mode optical modules and one multi-mode optical module, 2 HDMI interfaces, and 2 network interfaces, which can adapt to the input of multiple data formats and select the format of data output. It contains an Atlas200 acceleration module, which can deploy the yolov8 neural network algorithm to run target recognition. It also contains 64G-emmc and 256G-SD cards for on-board data storage.

[0058] The target tracking and identification circuit module solution integrates advanced computing and processing, stable and reliable power supply and control, efficient and flexible storage and caching, and diversified receiving and sending interfaces. It can realize real-time tracking and identification of targets in images and meet the needs of various application scenarios.

[0059] Example 3

[0060] According to the attached Figure 2 As shown, the specific steps are as follows:

[0061] Step 1: Select the corresponding interface circuit according to the specific input data format:

[0062] The cameralink format data input is decoded into the image's horizontal synchronization signal, field synchronization signal, and data signal through the cameralink in / out IP module;

[0063] The data input in sdi format is decoded into the line synchronization, field synchronization and data signals of the image through the SDI_3G in / out IP module;

[0064] The image data transmitted by optical fiber is decoded into the line synchronization, field synchronization and data signals of the image by setting the corresponding parameters through the 8b10b IP module;

[0065] The image data in HDMI format is decoded into the line synchronization, field synchronization, and RGB data signals through the HDMI IN IP, and then the RGB data is converted into the YUV data format for use;

[0066] The yuv data obtained by decoding and outputting the above-mentioned various types of data are manually selected (the selection interface instruction is sent from the control end through the serial port), and the selected data is prepared for target tracking and identification processing.

[0067] Step 2: The data is stored in the specified location of the PS end DDR in the form of a data stream through the Video in to AXI4-Stream IP and VDMA IP. The PS end reads the cache area image and obtains the position offset information through DCF algorithm calculation.

[0068] Step 3: The data is stored in the specified location of the PL-side DDR through the fdma_dbuf IP and FDMA IP. The Atlas200 device communicates with the SOC through PCIE and reads the image data at the specified location of the PL-side DDR through the XDMA IP AXI bus port. After receiving the data, Atlas performs the YOLOV8 neural network algorithm and returns the identification ID and location coordinate information of the algorithm operation through PCIE.

[0069] Step 4: Send the obtained position offset information to the servo control system through the serial port.

[0070] Step 5: Superimpose the obtained identification ID and location coordinate information with the cached image data read from the DDR of the PL end through the FDMA IP, and then convert the video data into the corresponding output format according to the needs of the output end.

[0071] Step 6: The data output in Cameralink format is encoded by the cameralink in / out IP module and converted into 4 pairs of differential data and 1 pair of differential clocks;

[0072] The data output in sdi format is encoded by the SDI_3G in / out IP module and converted to SDI data format for output; the image data transmitted by optical fiber is encoded by the 8b10b IP module and output through optical fiber;

[0073] HDMI format output, first convert the yuv data into RGB format data through format conversion, and then output 3 pairs of differential data and 1 pair of differential clocks through HDMIOUTIP.

[0074] Step 7. The position offset information finally output will guide the servo system to rotate, control the target to be within the lens range, and the display end will display the target identification ID and identification coordinate frame.

[0075] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A target tracking and identification circuit based on SOC, characterized in that: include: A computing and processing unit, the computing and processing unit includes a 7045-SOC chip and an Atlas200 acceleration module. The 7045-SOC chip is internally provided with a cameralink in / out IP module, an SDI_3G in / out IP module and an 8b10b IP module. The 7045-SOC chip cooperates with the Atlas200 acceleration module to realize the functions of data encoding and decoding, target tracking, recognition calculation, image character data superposition, and data encoding display output of the entire system; Power supply and control unit to achieve stable and sequential power supply of circuit boards; Storage and cache unit, to realize data cache and storage requirements during operation; The receiving and sending unit is used to realize the image data receiving and sending requirements during operation; The receiving and sending unit includes a cameralink transceiver connector, an SDI transceiver connector, a multi-channel MT optical fiber interface, an HDMI interface socket and a network interface socket.

2. The target tracking and identification circuit based on SOC according to claim 1, characterized in that: The cameralink in / out IP module, SDI_3G in / out IP module and 8b10bIP module exchange data with the outside world through the cameralink transceiver connector, the SDI transceiver connector and the multi-channel MT optical fiber interface respectively.

3. The target tracking and identification circuit based on SOC according to claim 1, characterized in that: The HDMI interface socket is used to receive and transmit high-definition digital video and audio signals, and the network interface socket is used to achieve the connection of network signals.

4. The target tracking and identification circuit based on SOC according to claim 1, characterized in that: The multi-channel MT optical fiber interface is divided into two single-mode optical modules and one multi-mode optical module, the cameralink transceiver connector is provided with 4 groups of Cameralink external transceiver interfaces, the SDI transceiver connector is provided with 4 SDI receptions, the HDMI interface socket is provided with 2, and the network interface socket is provided with 2.

5. The target tracking and identification circuit based on SOC according to claim 1, characterized in that: The Atlas200 acceleration module is physically connected to the 7045-SOC chip via PCIE.

6. The target tracking and identification circuit based on SOC according to claim 1, characterized in that: The Atlas200 acceleration module can deploy the yolov8 neural network algorithm to run target recognition.

7. The target tracking and identification circuit based on SOC according to claim 1, characterized in that: The power supply and control unit includes a 4644 power supply module, a 51200 power supply module and a 706 monitoring module.

8. The target tracking and identification circuit based on SOC according to claim 1, characterized in that: The storage and cache unit includes a DDR3 cache chip, an EMMC storage chip, an SD card memory and an external USB storage interface.

9. The target tracking and identification circuit based on SOC according to claim 8, characterized in that: The EMMC storage chip and SD card memory use 64G-emmc and 256G-SD card respectively for on-board data storage.

Images