Ballistic calculation system and method

Abstract

The invention relates to a ballistic calculation system and method, which are applied to the field of image processing, and the ballistic calculation system comprises an environmental parameter acquisition unit, an image data acquisition unit, an FPGA unit and an embedded platform. The environmental parameter acquisition unit is used for acquiring environmental parameters of an external environment; the image data acquisition unit is used for acquiring image data of a target; the FPGA unit is in communication connection with the environmental parameter acquisition unit and the image data acquisition unit, and is used for providing an interface in communication connection with the environmental parameter acquisition unit, receiving the image data and providing a power supply for the trajectory resolving system; and the embedded platform is in communication connection with the FPGA unit and is used for carrying out target identification according to the image data and carrying out target tracking and trajectory calculation according to the identified target and the environment parameters. The method has the technical effects that the target recognition accuracy is improved, and the intelligent level of the system is improved.

Description

Technical Field

[0001] This application relates to the field of image processing technology, and in particular, to a ballistic calculation system and method. Background Art

[0002] In the digital and intelligent era, with the continuous development of advanced information technology, intelligent warfare has also become one of the important combat methods in modern warfare. The rapid popularization of digital technology has greatly improved the level of military informatization. The wide application of various sensors, communication devices, and information processing systems has provided the military with a large amount of real-time data, realizing the all-round perception of the battlefield.

[0003] When hitting a target and performing ballistic calculation, it is necessary to first obtain the image of the target to be hit for target recognition and target tracking, and then perform ballistic calculation to obtain the offsets in the horizontal and vertical directions. Existing ballistic calculation systems usually separately collect video or image information through different detection optical paths for separate target recognition, and perform ballistic calculation after identifying the target. However, the method of separately collecting images and identifying targets through different channels is difficult to accurately identify targets and perform target tracking, and the recognition accuracy is low, which may lead to a low accuracy of ballistic calculation. Summary of the Invention

[0004] To help solve the problem that the method of separately collecting images and identifying targets through different channels is difficult to accurately identify targets and perform target tracking, resulting in a low accuracy of ballistic calculation, this application provides a ballistic calculation system and method.

[0005] In a first aspect, this application provides a ballistic calculation system, adopting the following technical solution: The ballistic calculation system includes an environmental parameter acquisition unit, an image data acquisition unit, an FPGA unit, and an embedded platform;

[0006] The environmental parameter acquisition unit is used to acquire the environmental parameters of the external environment;

[0007] The image data acquisition unit is used to acquire the image data of the target;

[0008] The FPGA unit is communicatively connected to the environmental parameter acquisition unit and the image data acquisition unit, and is used to provide an interface for communicatively connecting to the environmental parameter acquisition unit, receive the image data, and provide power for the ballistic calculation system;

[0009] The embedded platform is communicatively connected to the FPGA unit, and is used to perform target recognition based on the image data, and perform target tracking and ballistic calculation based on the recognized target and the environmental parameters.

[0010] In a specific feasible implementation, the ballistic calculation system further includes a display unit, and the display unit is used to display the environmental parameters in real time.

[0011] In a specific feasible implementation, the environmental parameter acquisition unit includes a laser ranging module, an attitude sensor module, a temperature and humidity sensor module, and a barometric pressure sensor module;

[0012] The laser ranging module is used to measure the target distance;

[0013] The attitude sensor module is used to detect the attitude data of the firearm;

[0014] The temperature and humidity sensor module is used to measure the temperature and humidity of the external environment;

[0015] The barometric pressure sensor module is used to measure the barometric pressure of the external environment.

[0016] In a specific feasible implementation, the image data acquisition unit includes an infrared detection module and a visible light detection module;

[0017] The infrared detection module is used to obtain the infrared image of the target;

[0018] The visible light detection module is used to obtain the visible light image of the target.

[0019] In a specific feasible implementation, the model of the FPGA unit is JFMK50T.

[0020] In a specific feasible implementation, the embedded platform is an embedded SOC platform, and the model is Hisilicon 3559AV100.

[0021] In a second aspect, the present application provides a ballistic calculation method, adopting the following technical solution: The method includes:

[0022] Obtain the infrared image and the visible light image of the target to be struck;

[0023] Perform target recognition based on the infrared image and the visible light image, and select and lock the target to be tracked;

[0024] Use the look-up table method to calculate the first horizontal offset and the first vertical offset of the ballistic trajectory for striking the target to be tracked;

[0025] Continuously track the target to be tracked, input the current motion data of the target to be tracked into a preset trajectory prediction model, and output the trajectory prediction horizontal offset and the trajectory prediction vertical offset;

[0026] Combine the first horizontal offset and the trajectory prediction horizontal offset to generate a ballistic horizontal offset, and combine the first vertical offset and the trajectory prediction vertical offset to generate a ballistic vertical offset.

[0027] In a specific implementable embodiment, the target recognition based on the infrared image and the visible light image includes:

[0028] Perform denoising processing on the infrared image and the visible light image respectively to generate a denoised infrared image and a denoised visible light image;

[0029] Extract features from the denoised infrared image and the denoised visible light image respectively, and generate target infrared features and target visible light features;

[0030] Perform feature fusion on the target infrared features and the target visible light features according to a preset weight, and generate target fusion features;

[0031] Input the target fusion features into a preset recognition model for inference calculation to identify the target to be struck.

[0032] In a specific implementable embodiment, the calculation of the first horizontal offset and the first vertical offset of the ballistic trajectory for striking the target to be tracked by using the look-up table method includes:

[0033] Obtain the environmental parameters of the external environment, where the environmental parameters include the target distance, the temperature and humidity of the external environment, the air pressure of the external environment, and the attitude data of the firearm;

[0034] Look up a preset ballistic table according to the environmental parameters to obtain the first horizontal offset and the first vertical offset of the ballistic trajectory for striking the target to be tracked.

[0035] In a specific implementable embodiment, the preset trajectory prediction model is a trajectory prediction model constructed based on the historical data of target tracking.

[0036] In summary, the present application has the following beneficial technical effects:

[0037] Obtain an infrared image and a visible light image through dual-channel detection. By combining the two spectra, the accuracy of target recognition and tracking can be improved. The ballistic solution system of the present application combines an FPGA platform and an embedded SOC platform, integrates multiple functions such as environmental parameter acquisition, target detection, target recognition, target tracking, ballistic solution, and threat warning. It has a high degree of integration and a small volume, and can be applied to a variety of individual soldier equipment, improving the intelligent level of individual soldier equipment. At the same time, it can effectively improve the combat personnel's perception ability of special targets and the environment, improve the user's combat decision-making ability, and effectively improve the combat success rate. Brief Description of the Drawings

[0038] Figure 1 It is the framework diagram of the ballistic calculation system in the embodiment of the present application;

[0039] Figure 2 It is the schematic diagram of the connection of each module of the ballistic calculation system in the embodiment of the present application;

[0040] Figure 3 It is the flow chart of the ballistic calculation method in the embodiment of the present application;

[0041] Figure 4 It is the schematic diagram of the ballistic calculation process in the embodiment of the present application;

[0042] Figure 5 It is the flow chart of the target recognition method in the embodiment of the present application. Specific embodiments

[0043] The following will be combined with Figures 1 - 5 to further elaborate on the present application in detail.

[0044] The embodiment of the present application discloses a ballistic calculation system. By using this system, intelligent target recognition and tracking can be realized, and intelligent ballistic calculation can be carried out. Through dual-channel video input, multi-spectral target feature fusion recognition can be achieved, improving the accuracy of target recognition, and further improving the accuracy of the ballistic calculation result.

[0045] In the digital and intelligent era, with the continuous development of advanced information technology, intelligent warfare has become one of the important warfare methods in modern warfare. The rapid popularization of digital technology has greatly improved the level of military informatization. The wide application of various sensors, communication devices, and information processing systems has provided the military with a large amount of real-time data, realizing the all-round perception of the battlefield. The rise of advanced technologies such as artificial intelligence and big data analysis has also injected powerful impetus into intelligent warfare, making command and decision-making more intelligent and efficient. As a strategic intelligent recognition system based on cutting-edge technologies, the dual-channel intelligent recognition system aims to achieve all-weather and all-terrain intelligence collection and combat command, better meeting the urgent needs of modern warfare for precision strikes and real-time intelligence.

[0046] When calculating the ballistic trajectory for a target to be struck, it is necessary to first obtain an image of the target to be struck for target recognition and target tracking, and then perform ballistic trajectory calculation to obtain the offsets in the horizontal and vertical directions. However, in existing ballistic trajectory calculation systems, the day-night observation and aiming systems used in multi-optical paths adopt separate-channel recognition for identification. The accuracy and variety of intelligent recognition are relatively low. The accuracy rate of single-channel target recognition is low, and it cannot achieve the effect of mutually assisting and improving the recognition accuracy between different spectra, which may lead to a decrease in the accuracy rate of ballistic trajectory calculation. In addition, existing ballistic trajectory calculation systems have a low level of intelligence and few functions; the types that can be recognized are single, and it is difficult to accurately recognize special clothing and special vehicles such as armored vehicles; most existing intelligent processing systems use foreign devices, and the level of localization is low; and it is difficult to solve the problem of dynamic target calculation. To help improve the accuracy rate of ballistic trajectory calculation results, this application provides a ballistic trajectory calculation system.

[0047] Referring to Figure 1 , the ballistic trajectory calculation system includes an environmental parameter acquisition unit 100, an image data acquisition unit 200, an FPGA unit 300, an embedded platform 400, and a display unit 500.

[0048] The environmental parameter acquisition unit 100 is used to acquire environmental parameters of the external environment. The image data acquisition unit 200 is used to acquire image data of the target. The FPGA unit 300 is communicatively connected to the environmental parameter acquisition unit 100 and the image data acquisition unit 200, and is used to provide an interface for communicatively connecting with the environmental parameter acquisition unit, receive image data, and provide power for the ballistic trajectory calculation system; wherein, the model of the FPGA unit is JFMK50T. The embedded platform 400 is communicatively connected to the FPGA unit 300, and is used to perform target recognition based on the image data, and perform target tracking and ballistic trajectory calculation based on the recognized target and environmental parameters; wherein, the embedded platform is an embedded SOC platform, and the model is Hisilicon 3559AV100. The display unit 500 is used to display the environmental parameters in real time.

[0049] The embodiments of this application use an embedded SOC platform combined with an FPGA platform to achieve target recognition, target tracking, and ballistic calculation. Among them, the model of the embedded SOC platform is the Hisilicon 3559AV100 platform. The Hisilicon 3559AV100 platform is a dual-core ARM Cortex A73 and a dual-core ARM Cortex A53 CPU, and is equipped with a dual-core NNIE neural network acceleration engine that can provide strong computing power support for intelligent recognition. In the interface board, an FPGA is used as an interface expansion platform to achieve the communication control of multiple sensors and the preprocessing of videos and images. The domestic FPGA platform used in the embodiments of this application is the JFMK50T of Fudan Microelectronics. Based on the XC7A50T as the main processor, it has rich IO interfaces and can expand the control of various sensors such as temperature, pressure, attitude, and laser rangefinder to obtain environmental parameter data. The combination of the embedded SOC platform and the powerful parallel processing ability of the FPGA platform can quickly preprocess the dual-channel image data and improve the real-time performance of the overall system.

[0050] Referring to Figure 2 , the environmental parameter acquisition unit 100 includes a laser ranging module 101, an attitude sensor module 102, a temperature and humidity sensor module 103, and a pressure sensor module 104. The laser ranging module 101 is used to measure the target distance; the attitude sensor module 102 is used to detect the attitude data of the firearm; the temperature and humidity sensor module 103 is used to measure the temperature and humidity of the external environment; the pressure sensor module 104 is used to measure the air pressure of the external environment. By integrating multiple sensors on the FPGA platform, comprehensive environmental information can be perceived, including temperature and humidity, air pressure, target distance, firearm attitude, etc.; in addition, the battlefield environmental parameters can be recorded in the form of taking pictures and videos, which can play a role in battlefield data analysis and data sharing, and improve the combat decision-making ability.

[0051] Referring to Figure 2 , the image data acquisition unit 200 includes an infrared detection module 201 and a visible light detection module 202. The infrared detection module 201 is used to obtain the infrared image of the target; the visible light detection module 202 is used to obtain the visible light image of the target. In the embodiments of this application, the detection system uses an infrared detector and a visible light detector, which can achieve day and night detection and realize the real-time acquisition of target images through multiple optical paths. The target features of different spectra are different, and different spectra can be used in different application scenarios; for example, in summer during the day, the infrared detection ability is weaker than that of the visible light path, and the visible light image features are more obvious, while at night, the infrared image features of the target personnel are more obvious. The ballistic calculation system of this application can achieve target recognition and target tracking. It uses the detection of dual optical path channels of infrared and visible light, fuses the features of different spectra, and mutually assists between different spectra in different environments and scenarios to improve the accuracy of target recognition.

[0052] It should be noted that in the target recognition and target tracking system of this application, when training the AI model, a dataset of special personnel and vehicles is used and labeled, the threat level is set, special targets are recognized and prominent warnings are given; special training for special personnel and vehicle targets is added. After the target is recognized in the infrared detection channel, the color characteristics of the visible light path are combined to further classify the target, identify its danger level, and give threat warnings; specifically, after recognition, the threat warning function is enabled, the threat level of the target can be recognized, and personnel targets with different threat levels are distinguished by different colors. Personnel with a high threat level are displayed in red and marked with a special alarm, personnel with a medium threat level are displayed in yellow, and personnel with a low threat level are displayed in green. At the same time, for small target personnel and vehicle targets, pixel filling is performed in the form of local image magnification, and recognition is performed after magnification to enhance the recognition effect of small targets, so that global and locally magnified recognition can be performed on small targets. Global magnification may lose some environmental information, but it is helpful for global detection; local magnification magnifies and recognizes local targets while ensuring environmental information, which can effectively improve the classification and recognition effect of targets. The combination of local magnification and global magnification improves the accuracy of recognition and tracking.

[0053] The display unit 500 can display the measured environmental parameters in real time, and can provide quick operations for common functions such as switching the optical path and enabling recognition, which is convenient for operators to operate and improves the combat efficiency.

[0054] In the solution of this application, infrared images and visible light images are obtained through dual-channel detection, and the accuracy of target recognition and tracking can be improved by combining the two spectra. The ballistic calculation system of this application combines the FPGA platform and the embedded SOC platform, integrates multiple functions such as environmental parameter acquisition, target detection, target recognition, target tracking, ballistic calculation, and threat warning, has a high degree of integration, and is small in size and can be applied to a variety of individual soldier equipment, improving the intelligent level of individual soldier equipment. At the same time, it can effectively improve the combat personnel's perception ability of special targets and the environment, improve the user's combat decision-making ability, and effectively improve the combat success rate.

[0055] Based on the above system, the embodiment of this application also discloses a ballistic calculation method.

[0056] Refer to Figure 3 , the method includes the following steps:

[0057] S10, obtain the infrared image and visible light image of the target to be struck.

[0058] Specifically, the infrared detection module and the visible light detection module in the image data acquisition unit are used to obtain the infrared image and the visible light image of the target to be struck respectively. Different spectral target features are different, and different spectra can be used for different application scenarios. Based on the combination of the two spectra, the accuracy of target recognition can be improved.

[0059] S20. Perform target recognition based on the infrared image and the visible light image, and select and lock the target to be tracked.

[0060] Specifically, based on the fusion features of the two spectral images, the target to be struck is accurately recognized; the FPGA platform receives the originally acquired infrared image and visible light image, and then transmits the preprocessed two spectral images to the embedded SOC platform for target recognition and target tracking. After the target is recognized, the target that the user needs to track is selected from the results of target recognition, and the target to be tracked is locked. After locking the target, only the target box of the tracked target is retained, and the target tracking algorithm is called for continuous target tracking.

[0061] S30. Calculate the first horizontal offset and the first vertical offset of the ballistic trajectory for striking the target to be tracked by using the look-up table method.

[0062] Specifically, after the embedded SOC platform recognizes and locks the target, for the target to be tracked that is locked and tracked, the look-up table method is used to calculate the first horizontal offset and the first vertical offset of the ballistic trajectory for striking the target. Specifically, referring to Figure 4 , after the target is locked, the environmental parameters of the external environment are obtained, and the environmental parameters include the target distance, the temperature and humidity of the external environment, the air pressure of the external environment, and the attitude data of the firearm; according to the environmental parameters, a preset ballistic table is searched, and the first horizontal offset (X0) and the first vertical offset (Y0) of the ballistic trajectory for striking the target to be tracked can be solved and obtained.

[0063] S40. Continuously track the target to be tracked, input the current motion data of the target to be tracked into a preset trajectory prediction model, and output the trajectory prediction horizontal offset and the trajectory prediction vertical offset.

[0064] Specifically, referring to Figure 4 , the target to be tracked will move continuously. After the target is locked, the locked target to be tracked is continuously tracked, and the current motion data of the target to be tracked is input into the preset trajectory prediction model for model inference, and the trajectory prediction horizontal offset (X1) and the trajectory prediction vertical offset (Y1) can be inferred and calculated and output. Among them, the preset trajectory prediction model is a trajectory prediction model constructed by the user in advance according to the historical data of previous target tracking. The trajectory prediction model is constructed through the historical tracking data, so as to judge the possible motion trajectory of the target to be tracked, and the trajectory prediction horizontal offset and the trajectory prediction vertical offset are predicted through the model.

[0065] S50, combining the first horizontal offset and the predicted trajectory horizontal offset to generate a trajectory horizontal offset, and combining the first vertical offset and the predicted trajectory vertical offset to generate a trajectory vertical offset.

[0066] Specifically, the first horizontal offset and vertical offset (X0, Y0) are obtained by table lookup, and the trajectory prediction horizontal offset and trajectory prediction vertical offset (X1, Y1) are obtained by large model inference calculation, and the two parts of the offset are fused and calculated to obtain the final trajectory horizontal offset and trajectory vertical offset (X, Y). Among them, the two parts of the offset fusion calculation method can be weighted calculation, different parts are assigned different weights and then fused, or it can be directly superimposed, or other calculation methods can be used. There is no restriction here, and users can choose according to actual conditions.

[0067] In this application, intelligent target identification and historical trajectory tracking are used to perform pre-judgment analysis, establish a ballistic mathematical model, and form corresponding ballistic strike guidance points. The intelligent ballistic solution function can make full use of environmental parameters for table lookup, as well as multi-dimensional information such as target motion trajectory for trajectory solution, which can effectively improve the product's strike accuracy and effectively solve the problem of dynamic target strikes.

[0068] In one embodiment, referring to Figure 5 , the target recognition method based on infrared images and visible light images can be specifically implemented as follows:

[0069] First, after receiving the two spectral images, the FPGA platform performs denoising preprocessing on the infrared image and the visible light image respectively, and generates infrared denoised images and visible light denoised images to improve the image quality and avoid the influence of noise on the target features as much as possible. After that, the FPGA platform transmits the denoising preprocessed images to the SOC platform, where the infrared denoised images and the visible light denoised images are extracted and the target infrared features and target visible light features are generated; then, the target infrared features and target visible light features are fused according to the preset weights, and the target fusion features are generated. The specific weight coefficients can be set by the user according to the actual situation, and there is no restriction here. Finally, the target fusion features are input into the preset recognition model for inference calculation to identify the target to be hit.

[0070] In the present application, the features extracted from the two spectral images are fused, which reduces the amount of recognition calculations while integrating the features of another optical path, thereby improving recognition accuracy and helping to improve the accuracy of subsequent trajectory solution results.

[0071] Figure 3It is a schematic flowchart of a ballistic calculation method in an embodiment. It should be understood that although Figure 3 each step in the flowchart is shown sequentially according to the indication of the arrow, these steps are not necessarily executed sequentially in the order indicated by the arrow; unless there is a clear description in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders; and Figure 3 at least a part of the steps in can include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed alternately or alternately with at least a part of other steps or sub-steps or stages of other steps.

[0072] This specific embodiment is only an explanation of the present invention, and it is not a limitation of the present invention. Those skilled in the art can make modifications to this embodiment without creative contributions according to needs after reading this specification, but as long as it is within the scope of the claims of the present invention, it is protected by the patent law.

Claims

1. A trajectory calculation system, characterized in that: The trajectory calculation system includes an environmental parameter acquisition unit, an image data acquisition unit, an FPGA unit and an embedded platform; The environmental parameter acquisition unit is used to acquire environmental parameters of the external environment; The image data acquisition unit is used to acquire image data of the target; The FPGA unit is in communication with the environmental parameter acquisition unit and the image data acquisition unit, and is used to provide an interface in communication with the environmental parameter acquisition unit, receive the image data, and provide power for the trajectory solution system; The embedded platform is communicatively connected with the FPGA unit, and is used for performing target recognition according to the image data, and performing target tracking and trajectory calculation according to the recognized target and the environmental parameters.

2. The trajectory calculation system according to claim 1, characterized in that: The trajectory calculation system further includes a display unit, and the display unit is used to display the environmental parameters in real time.

3. The trajectory calculation system according to claim 1, characterized in that: The environmental parameter acquisition unit includes a laser ranging module, a posture sensor module, a temperature and humidity sensor module, and an air pressure sensor module; The laser ranging module is used to measure the target distance; The posture sensor module is used to detect the posture data of the firearm; The temperature and humidity sensor module is used to measure the temperature and humidity of the external environment; The air pressure sensor module is used to measure the air pressure of the external environment.

4. The trajectory calculation system according to claim 1, characterized in that: The image data acquisition unit includes an infrared detection module and a visible light detection module; The infrared detection module is used to obtain an infrared image of the target; The visible light detection module is used to obtain a visible light image of the target.

5. The trajectory calculation system according to claim 1, characterized in that: The model of the FPGA unit is JFMK50T.

6. The trajectory calculation system according to claim 1, characterized in that: The embedded platform is an embedded SOC platform, model is HiSilicon 3559AV100.

7. A trajectory calculation method, characterized in that: The method comprises: Acquire infrared images and visible light images of the target to be hit; Performing target recognition according to the infrared image and the visible light image, and selecting and locking a target to be tracked; Using a table lookup method to calculate a first horizontal offset and a first vertical offset of a trajectory of the target to be tracked; Continuously tracking the target to be tracked, inputting current motion data of the target to be tracked into a preset trajectory prediction model, and outputting a predicted trajectory horizontal offset and a predicted trajectory vertical offset; The first horizontal offset and the predicted trajectory horizontal offset are combined to generate a trajectory horizontal offset, and the first vertical offset and the predicted trajectory vertical offset are combined to generate a trajectory vertical offset.

8. The trajectory calculation method according to claim 7, characterized in that: The performing target recognition according to the infrared image and the visible light image comprises: Performing denoising processing on the infrared image and the visible light image respectively to generate an infrared denoised image and a visible light denoised image; Extracting features from the infrared denoised image and the visible light denoised image respectively, and generating target infrared features and target visible light features; The target infrared feature and the target visible light feature are fused according to a preset weight, and a target fusion feature is generated; The target fusion features are input into a preset recognition model for inference calculation to identify the target to be hit.

9. The trajectory calculation method according to claim 7, characterized in that: The method of using a table lookup method to calculate the first horizontal offset and the first vertical offset of the trajectory of the target to be tracked comprises: Acquiring environmental parameters of the external environment, wherein the environmental parameters include target distance, temperature and humidity of the external environment, air pressure of the external environment, and posture data of the firearm; A preset trajectory table is searched according to the environmental parameters to obtain a first horizontal offset and a first vertical offset of the trajectory for striking the target to be tracked.

10. The trajectory calculation method according to claim 7, characterized in that: The preset trajectory prediction model is a trajectory prediction model constructed based on historical data of target tracking.

Images