Aerial target tracking method, device, storage medium, and controller

Abstract

The present invention relates to the field of aerial tracking technology, and discloses an aerial target tracking method, device, storage medium, and controller. The method comprises processing a video frame collected by a guide head according to a target search range to obtain a target image, wherein the target search range is determined by an initialization target search center or historical target state information; performing target detection on the target image to obtain a detection target; tracking the detection target in the target image according to the target search range to obtain a tracking position of the detection target; and obtaining target state information of the detection target according to the tracking position of the detection target and historical target state information. The method processes the video frame collected by the guide head according to the target search range, performs target detection and tracking on the target image obtained by the processing, and realizes the detection and tracking of full-scale targets.

Description

Technical Field

[0001] The present invention relates to the field of aerial tracking technology, and in particular to an aerial target tracking method, device, storage medium and controller. Background Art

[0002] Throughout the combat mission, the image seeker needs to be able to adapt to full-scale target detection and tracking.

[0003] Related high-precision detection methods, such as the YOLO neural network algorithm, achieve multi-scale detection by setting up multiple detection heads. However, considering resource issues such as computing power and storage, the number of detection heads is limited, and the scale of multi-scale also has a certain range, which cannot achieve full-scale target detection. How to achieve full-scale target detection and tracking is a technical problem that technicians in this field need to solve. Summary of the invention

[0004] The present invention aims to solve one of the technical problems in the related art at least to a certain extent. To this end, one object of the present invention is to propose an aerial target tracking method to achieve detection and tracking of full-scale targets.

[0005] The second objective of the present invention is to provide an aerial target tracking device.

[0006] A third object of the present invention is to provide a computer-readable storage medium.

[0007] A fourth object of the present invention is to provide a controller.

[0008] To achieve the above-mentioned purpose, the first aspect of the present invention proposes an aerial target tracking method, which includes: processing the video frames captured by the guidance head according to the target search range to obtain the target image, wherein the target search range is determined by the initialization target search center or the historical target status information; performing target detection on the target image to obtain the detected target; tracking the detected target in the target image according to the target search range to obtain the tracking position of the detected target; and obtaining the target status information of the detected target according to the tracking position of the detected target and the historical target status information.

[0009] According to the aerial target tracking method of the embodiment of the present invention, the video frames collected by the guidance head are processed according to the target search range, and the processed target image is subjected to target detection to detect all targets. The detection target in the target image is tracked according to the target search range, the tracking position of the detection target is determined, and the target state information of the detection target is determined according to the tracking position of the detection target and the historical target state information, so as to realize the detection and tracking of full-scale targets.

[0010] In addition, the aerial target tracking method proposed in the above embodiment of the present invention may also have the following additional technical features:

[0011] According to one embodiment of the present invention, the target image includes at least one of a first image, a second image and a third image, and the video frame captured by the guide head is processed according to the target search range to obtain the target image, including: if the target search range is smaller than the lower limit of a preset size range, then a screenshot of the video frame is taken with the center of the target search range as the center according to a preset image size to obtain the first image; if the target search range is within the preset size range, then the video frame is compressed according to the preset image size to obtain the second image; if the target search range is larger than the upper limit of the preset size range, then the video frame is edge-filled according to the preset image size and then compressed to obtain the third image.

[0012] According to one embodiment of the present invention, performing target detection on the target image to obtain the detected target includes: if the target image is the first image, performing target detection after feature extraction processing on the target image to obtain a first target; if the target image is the second image or the third image, performing target detection on the target image using a target detection algorithm to obtain a second target; and fusing the first target and / or the second target to obtain the detected target.

[0013] According to one embodiment of the present invention, performing target detection after feature extraction processing on the target image to obtain a first target includes: using a preset convolution kernel to extract features from the target image to obtain an abstract target image, wherein the size of the preset convolution kernel is determined by the target scale range; counting the number of outer circle pixel points whose outer circle pixel value of the abstract target in the abstract target image is greater than a preset gradient threshold minus the mean value of the inner circle pixels; and recording the abstract target whose outer circle pixel point number is greater than the preset point count threshold as the first target.

[0014] According to one embodiment of the present invention, tracking the detection target according to the target search range includes: tracking and searching the detection target within the target search range, and calculating a tracking response value of the detection target; if the tracking response value is greater than a preset response threshold, extracting the center position of the detection target; if the tracking response value is less than the preset response threshold, updating the target search range according to the detection target and the historical target status information, and executing the step of tracking and searching the detection target within the target search range.

[0015] According to one embodiment of the present invention, the target state information includes a target position and target state information, and the target state information of the detection target is obtained based on the tracking position of the detection target and the historical target state information, including: inferring the position of the detection target based on the historical target position and historical state information in the historical target state information to determine the inferred position; when the tracking position of the detection target is within a preset range of the inferred position, recording the tracking position as the target position of the detection target; and calculating the target state information of the detection target based on the tracking position and the historical target position and historical state information in the historical target state information.

[0016] According to an embodiment of the present invention, the method further includes: storing the target state information, and recording the target state information as the historical target state information.

[0017] To achieve the above-mentioned purpose, the second aspect of the embodiment of the present invention proposes an aerial target tracking device, which includes: a pre-processing module, which is used to process the video frames collected by the guide head according to the target search range to obtain a target image, wherein the target search range is determined by the initialization target search center or the historical target status information; a detection module, which is used to perform target detection on the target image to obtain a detected target; a tracking module, which is used to track the detected target in the target image according to the target search range to obtain the tracking position of the detected target; and a post-processing module, which is used to obtain the target status information of the detected target according to the tracking position of the detected target and the historical target status information.

[0018] To achieve the above-mentioned purpose, the third aspect of the present invention proposes a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the aerial target tracking method proposed in the first aspect of the present invention is implemented.

[0019] To achieve the above-mentioned purpose, the fourth aspect of the present invention proposes a controller, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the aerial target tracking method proposed in the first aspect of the present invention is implemented.

[0020] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 is a flow chart of an aerial target tracking method according to an embodiment of the present invention;

[0022] Figure 2is a flow chart of determining a first target according to an embodiment of the present invention;

[0023] Figure 3 is a schematic diagram of a preset convolution kernel according to a specific embodiment of the present invention;

[0024] Figure 4 is a flow chart of determining target state information of a detection target according to an embodiment of the present invention;

[0025] Figure 5 is a schematic diagram of an aerial target tracking device according to an embodiment of the present invention;

[0026] Figure 6 4 is a structural block diagram of a controller according to an embodiment of the present invention. DETAILED DESCRIPTION

[0027] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0028] The following is attached to the instruction manual Figure 1-6 And specific implementation methods are described in detail for the aerial target tracking method, device, storage medium, and controller of the embodiments of the present invention.

[0029] Figure 1 FIG. 1 is a flow chart of an aerial target tracking method according to an embodiment of the present invention. Figure 1 As shown, the aerial target tracking method may include:

[0030] S101, processing the video frames collected by the guidance head according to the target search range to obtain a target image, wherein the target search range is determined by the initialization target search center or the historical target state information;

[0031] S102, performing target detection on the target image to obtain a detected target;

[0032] S103, tracking the detection target in the target image according to the target search range to obtain a tracking position of the detection target;

[0033] S104, obtaining target state information of the detection target according to the tracking position of the detection target and the historical target state information.

[0034] Specifically, the image acquisition module of the guide head can be used to collect video of the target aerial area. The video collected by the image acquisition module is received, and target detection and tracking are performed on each video frame to determine the target state information of the aerial target.

[0035] It should be noted that there may be one or more targets in the target aerial area, and the distances of each target from the guide head are different. The farther the missile-target distance (the distance between the guide head and the aerial target), the fewer pixels of the target in the video frame, and the closer the missile-target distance, the more pixels of the target in the video frame. As the missile-target distance gets closer, the target will continue to enlarge, and the search range of the target will also continue to expand. In order to detect full-size targets, the embodiment of the present invention processes the video frames collected by the guide head according to the target search range to obtain one or more target images.

[0036] When determining the target search range, the initial target search center of each target in the air can be determined based on the radar pointing information or the operator's operation information, and the target search range of each target can be determined by inference based on the initial target search center. The target search range of each target can also be determined by inference based on the historical target state information of each target in the air.

[0037] As an example, at the time of initial detection and tracking, if the target state information of each target in the air is not tracked, the initialization target search center of each target in the air can be determined according to the radar pointing information or the operation information of the operator, and then the target search range of each target can be determined by inference based on the initialization target search center of each target. After the target state information of each target in the air is tracked, the target state information of each target can be stored, and the target state information of each target can be recorded as historical target state information. The target search range of each target can be determined by inference based on the historical target state information of each target.

[0038] It should be noted that determining the target search range of each target by inferring the initial target search center of each target is not limited to the initial detection and tracking moment, but is also applicable to other situations.

[0039] To detect aerial targets at all scales, after determining the target search range of each target, the video frames are processed to obtain at least one target image containing the full-scale target. Target detection is performed on each target image to determine the detected targets of each size. The detected targets in each target image are tracked according to the target search range to determine the tracking position of each detected target. The target state information of each detected target is determined based on the tracking position of each detected target and the historical target state information.

[0040] In one embodiment of the present invention, the target image includes at least one of the first image, the second image and the third image, and the video frame captured by the guide head is processed according to the target search range to obtain the target image, which may include:

[0041] If the target search range is smaller than the lower limit of the preset size range, a screenshot of the video frame is taken with the center of the target search range as the center according to the preset image size to obtain a first image;

[0042] If the target search range is within the preset size range, compressing the video frame according to the preset image size to obtain a second image;

[0043] If the target search range is larger than the upper limit of the preset size range, the video frame is edge-filled according to the preset image size and then compressed to obtain a third image.

[0044] When the distance between the projectile and the target is far, the pixel size of the target in the image is low, perhaps only 2~3pix. As the distance between the projectile and the target gets closer, the pixel size of the target is enlarged until it fills the entire field of view. In the process of target detection and tracking, there are targets that are too large or too small.

[0045] Since the video frames captured by the seeker usually have a large field of view, the video frames must be compressed to a preset image size for target detection and target tracking. The number of detection heads in the relevant target detection network is limited, and the scale of multiple scales also has a certain range. In addition, the relevant target detection algorithm generally defines targets below 30pix as small targets. Targets that may only have 2~3pix in the video frame will be more difficult to detect.

[0046] The embodiment of the present invention is to identify full-scale targets. A preset size range is set. When the target search range is within the preset size range, after compressing the video frame into a target image of a preset image size, the relevant target detection network can identify the target within the target search range in the target image.

[0047] For targets with a long distance between the projectile and the target and a low pixel size, when the corresponding target search range is smaller than the lower limit of the preset size range, in order to prevent the detection of targets with too small size, the video frame is captured with the center of the target search range as the center according to the preset image size. While ensuring that the first image obtained by the screenshot can be further targeted and tracked, the tiny target features can be protected from compression.

[0048] For a target with a moderate missile-target distance, when the corresponding target search range is within a preset size range, the video frame can be directly compressed according to the preset size range.

[0049] For targets that are close to the target, the target search range is continuously enlarged. When the target search range is larger than the upper limit of the preset size range, the edge of the video frame can be filled and the filled video frame can be compressed according to the preset image size.

[0050] In the embodiment of the present invention, data enhancement may be further performed on the first image, the second image and the third image.

[0051] In one embodiment of the present invention, performing target detection on a target image to obtain a detected target may include:

[0052] If the target image is the first image, then feature extraction is performed on the target image and then target detection is performed to obtain the first target;

[0053] If the target image is the second image or the third image, a target detection algorithm is used to perform target detection on the target image to obtain a second target;

[0054] The first target and / or the second target are fused to obtain a detection target.

[0055] Specifically, when the target image is the first image, the target has only a few pixels in the first image, and the target features are few, so the relevant target detection network cannot detect it. In order to detect the small-sized target in the first image, the first image is subjected to feature extraction processing to abstract the target in the first image, and then target detection is performed on the abstracted first image to determine the small-sized first target.

[0056] When the target image is the second image or the third image, the target size in the second image or the third image is moderate. For the detection of intermediate-scale targets, the relevant detection algorithm has a very good effect. The target detection algorithm, such as the YOLO neural network and the SSD detection network, can be used to perform target detection on the target image to obtain the medium-sized second target.

[0057] When there is only a small-sized first target, one or more small-sized first targets are fused to obtain all the detected targets. When there is only a medium-sized second target, one or more medium-sized second targets are fused to obtain all the detected targets. When there are both first targets and second targets, one or more small-sized first targets and one or more medium-sized second targets are fused to obtain all the detected targets.

[0058] In one embodiment of the present invention, Figure 2 As shown, after feature extraction processing is performed on the target image, target detection is performed to obtain a first target, including:

[0059] S201, using a preset convolution kernel to extract features of a target image and abstract the target image, wherein the size of the preset convolution kernel is determined by the target scale range;

[0060] S202, counting the number of outer circle pixel points whose outer circle pixel value of the abstract target in the abstract target image minus the inner circle pixel mean is greater than a preset gradient threshold;

[0061] S203: Record the abstract target whose outer circle pixel count is greater than a preset pixel count threshold as the first target.

[0062] Specifically, for a target with only a few pixels in the first image, when determining the first target, a preset convolution kernel is used to perform feature extraction on the target image. The preset convolution kernel can extract the gradient of the surrounding and middle areas of the small-sized target, and abstract the small-sized target into a black middle area with white surrounding areas (abstract target), thereby enhancing the target with only a few pixels.

[0063] Calculate the mean value of the inner circle pixels of the abstract target in the abstract target image obtained after feature extraction. Count the number of outer circle pixels whose outer circle pixel value minus the mean value of the inner circle pixels is greater than the preset gradient threshold. If the number of outer circle pixels is greater than the preset point threshold, the abstract target is recorded as the first target. If the number of outer circle pixels is less than the preset point threshold, the abstract target is recorded as an interference. By comparing the number of outer circle pixels with the preset point threshold, interference filtering can be achieved.

[0064] The embodiments of the present invention solve the problem of full-scale target detection and can realize tiny target recognition and detection.

[0065] It should be noted that the preset convolution kernel in the embodiment of the present invention can be as follows: Figure 3 The 9*9 convolution kernel shown in the figure has the convolution meaning of the mean grayscale value of the outer circle pixels minus the mean grayscale value of the pixels in the middle 3*3 area. The middle area is set to 0. Therefore, the target scale extracted by the preset convolution kernel target can be scaled within a certain range. Figure 3 The convolution kernel size shown is only one type. In practical applications, the convolution kernel size can be designed accordingly according to the target scale range.

[0066] In one embodiment of the present invention, tracking the detection target according to the target search range includes:

[0067] Track and search the detection target within the target search range, and calculate the detection target tracking response value;

[0068] If the tracking response value is greater than the preset response threshold, the center position of the detection target is extracted;

[0069] If the tracking response value is less than the preset response threshold, the target search range is updated according to the detection target and the historical target state information, and the step of tracking and searching the detection target within the target search range is executed.

[0070] Specifically, to achieve designated tracking, a detection target within a target search range in a target image is tracked and searched, and a tracking response value corresponding to the detection target is calculated.

[0071] When the tracking response value is greater than the preset response threshold, it can be determined that the detection target is tracked. After it is determined that the detection target is tracked, the target center of the detection target is extracted, and the center position of the detection target is determined, thereby obtaining the tracking position of the detection target. After obtaining the tracking position of the detection target, the historical target position can be updated according to the current tracking position to adapt to the changing target.

[0072] When the tracking response value is less than the preset response threshold, it indicates that the target is lost. To track the detection target, the detection target is refreshed and updated according to the detected detection target and its corresponding target history status information, so as to further track and search the detection target within the target search range until the detection target is tracked. When it is determined that the detection target is tracked, the center position of the detection target is extracted to determine the center position of the detection target, thereby obtaining the tracking position of the detection target.

[0073] In one embodiment of the present invention, Figure 4 As shown, the target state information includes the target position and the target state information. According to the tracking position of the detected target and the historical target state information, the target state information of the detected target is obtained, including:

[0074] S301, inferring the position of the detection target according to the historical target position and the historical state information in the historical target state information to determine the inferred position;

[0075] S302, when the tracking position of the detection target is within a preset range of the inference position, recording the tracking position as the target position of the detection target;

[0076] S303, calculating and obtaining target state information of the detection target according to the tracking position and the historical target position and the historical state information in the historical target state information.

[0077] Specifically, the inference position of each detected target is determined based on the historical target state information corresponding to each detected target, such as the target motion curve and motion velocity vector. According to the inference position corresponding to the target detection, the rationality of the obtained tracking position is judged.

[0078] If the tracking position of the detection target is within the preset range of the reasoning position, it means that the tracking position is reasonable. According to the relevant reasoning curve, the target state information of the detection target can be calculated according to the tracking position of the detection target and the historical target position and historical state information in the corresponding historical target state information. If the tracking position of the detection target is not within the preset range of the reasoning position, it means that the tracking position is unreasonable. The target can be further tracked.

[0079] In one embodiment of the present invention, the aerial target tracking method further includes: storing the target state information, and recording the target state information as historical target state information.

[0080] Specifically, after the target state information of the detection target is obtained, the target state information of the detection target is managed, the target state information is stored in a register, and the historical target state information of the detection target in the register is further updated.

[0081] The aerial target tracking method of the embodiment of the present invention can realize long-distance target positioning and tracking to achieve precise strikes in aerial combat missions.

[0082] The invention provides an aerial target tracking device.

[0083] Figure 5 FIG. 1 is a schematic diagram of an aerial target tracking device according to an embodiment of the present invention. Figure 5 As shown, the aerial target tracking device 100 may include a pre-processing module 10 , a detection module 20 , a tracking module 30 and a post-processing module 40 .

[0084] Among them, the pre-processing module 10 is used to process the video frames collected by the guide head according to the target search range to obtain the target image, wherein the target search range is determined by the initialization target search center or the historical target state information; the detection module 20 is used to perform target detection on the target image to obtain the detected target; the tracking module 30 is used to track the detected target in the target image according to the target search range to obtain the tracking position of the detected target; the post-processing module 40 is used to obtain the target state information of the detected target according to the tracking position of the detected target and the historical target state information.

[0085] The pre-processing module 10 receives the video collected by the image acquisition module. Since there may be one or more targets in the target air area, the distances of each target from the guide head are different. The farther the projectile-target distance (the distance between the guide head and the air target), the fewer pixels of the target in the video frame, and the closer the projectile-target distance, the more pixels of the target in the video frame. As the projectile-target distance gets closer, the target will continue to enlarge, and the search range of the target will also continue to expand. In order to detect full-size targets, the pre-processing module 10 is used to infer and determine the target search range, and the video frames collected by the guide head are processed according to the target search range to obtain one or more target images.

[0086] Specifically, when the distance between the projectile and the target is far, the pixel size of the target in the image is low, perhaps only 2~3pix. As the distance between the projectile and the target gets closer, the pixel size of the target is enlarged until it fills the entire field of view. In the process of target detection and tracking, there are targets that are too large or too small.

[0087] Since the video frames captured by the seeker usually have a large field of view, the video frames must be compressed to a preset image size for target detection and target tracking. The number of detection heads in the relevant target detection network is limited, and the scale of multiple scales also has a certain range. In addition, the relevant target detection algorithm generally defines targets below 30pix as small targets. Targets that may only have 2~3pix in the video frame will be more difficult to detect.

[0088] The embodiment of the present invention is to identify full-scale targets. A preset size range is set. When the target search range is within the preset size range, after compressing the video frame into a target image of a preset image size, the relevant target detection network can identify the target within the target search range in the target image.

[0089] Specifically, for targets with a long distance between the projectile and the target and a low pixel size, when the corresponding target search range is smaller than the lower limit of the preset size range, in order to prevent the detection of targets with too small a size, a screenshot of the video frame is taken with the center of the target search range as the center according to the preset image size. While ensuring that the first image obtained by the screenshot can be further targeted and tracked, the tiny target features can also be protected from compression.

[0090] Specifically, for a target with a moderate missile-target distance, when its corresponding target search range is within a preset size range, the video frame can be directly compressed according to the preset size range.

[0091] For targets that are close to the target, the target search range is continuously enlarged. When the target search range is larger than the upper limit of the preset size range, the edge of the video frame can be filled and the filled video frame can be compressed according to the preset image size.

[0092] In the embodiment of the present invention, the pre-processing module 10 may further perform data enhancement on the first image, the second image and the third image.

[0093] In one embodiment of the present invention, the detection module 20 may include a point object detection unit, a mesoscale object detection unit, and a mid-scale object detection unit.

[0094] The point target detection unit mainly detects tiny targets of several pixels, and its main body is the target convolution extractor and interference filter.

[0095] When detecting a target with only a few pixels in the first image, the point target detection unit uses a preset convolution kernel to extract features of the target image. The preset convolution kernel can extract the gradient of the surrounding and middle areas of the small-sized target, abstract the small-sized target into an area with black in the middle and white around the periphery (abstract target), and enhance the target with only a few pixels. Calculate the mean value of the inner circle pixels of the abstract target in the abstract target image obtained after feature extraction. Count the number of outer circle pixels whose outer circle pixel value of the abstract target minus the mean value of the inner circle pixels is greater than the preset gradient threshold. If the number of outer circle pixel points is greater than the preset point threshold, the abstract target is recorded as the first target. If the number of outer circle pixel points is less than the preset point threshold, the abstract target is recorded as an interference. By comparing the number of outer circle pixel points with the preset point threshold, interference filtering can be achieved.

[0096] For the detection of intermediate-scale targets, the relevant detection algorithm has a very good effect. The medium-scale target detection unit can use conventional target detection methods, such as YOLO neural network and SSD detection network, to perform target detection on the target image and obtain a medium-sized second target.

[0097] The target detection fusion unit fuses the point target detection results and the mesoscale target detection results. The point target detection only extracts the point targets whose areas are smaller than the preset requirements, and the mesoscale target detection only extracts the targets whose areas are larger than the point target area threshold. The target detection results of the two detection modules are combined to obtain targets covering all scales.

[0098] The tracking module tracks and searches for the detection target within the target search range in the target image to achieve the specified tracking, and calculates the tracking response value corresponding to the detection target.

[0099] When the tracking response value is greater than the preset response threshold, it can be determined that the detection target is tracked. After it is determined that the detection target is tracked, the target center of the detection target is extracted, and the center position of the detection target is determined, thereby obtaining the tracking position of the detection target. After obtaining the tracking position of the detection target, the historical target position can be updated according to the current tracking position to adapt to the changing target.

[0100] When the tracking response value is less than the preset response threshold, it indicates that the target is lost. To track the detection target, the detection target is refreshed and updated according to the detected detection target and its corresponding target history status information, so as to further track and search the detection target within the target search range until the detection target is tracked. When it is determined that the detection target is tracked, the center position of the detection target is extracted to determine the center position of the detection target, thereby obtaining the tracking position of the detection target.

[0101] The post-processing module 40 performs reasoning based on the historical target state information corresponding to each detection target, such as the target motion curve and the motion velocity vector, to determine the reasoning position of each detection target. According to the reasoning position corresponding to the target detection, the rationality of the obtained tracking position is judged. If the tracking position of the detection target is within the preset range of the reasoning position, it means that the tracking position is reasonable. The target state information of the detection target can be calculated based on the relevant reasoning curve, the tracking position of the detection target and the historical target position and historical state information in the corresponding historical target state information. If the tracking position of the detection target is not within the preset range of the reasoning position, it means that the tracking position is unreasonable, and the target can be further tracked.

[0102] In one embodiment of the present invention, Figure 5 As shown, the aerial target tracking device 100 may also include a register. The register mainly stores the historical target state information of each detected target and the calculation information required by other modules.

[0103] It should be noted that, for other specific implementations of the aerial target tracking device provided in the embodiment of the present invention, reference may be made to other specific implementations of the aerial target tracking method in the above embodiment of the present invention.

[0104] The aerial target tracking device of the embodiment of the present invention can realize long-distance target positioning and tracking to achieve precise strikes in aerial combat missions.

[0105] The present invention provides a computer-readable storage medium.

[0106] In this embodiment, a computer program is stored on a computer-readable storage medium, and when the computer program is executed by a processor, the above-mentioned aerial target tracking method is implemented.

[0107] The invention provides a controller.

[0108] In this embodiment, the controller may include a memory and a processor. The memory stores a computer program. When the computer program is executed by the processor, the above-mentioned aerial target tracking method is implemented.

[0109] Figure 6 4 is a structural block diagram of a controller according to an embodiment of the present invention.

[0110] like Figure 6 As shown, the controller 500 includes: a processor 501 and a memory 503. The processor 501 and the memory 503 are connected, such as through a bus 502. Optionally, the controller 500 may also include a transceiver 504. It should be noted that in actual applications, the transceiver 504 is not limited to one, and the structure of the controller 500 does not constitute a limitation on the embodiments of the present invention.

[0111] The processor 501 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. It may implement or execute various exemplary logic blocks, modules and circuits described in conjunction with the disclosure of the present invention. The processor 501 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0112] The bus 502 may include a path to transmit information between the above components. The bus 502 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus 502 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 6 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0113] The memory 503 is used to store a computer program corresponding to the aerial target tracking method of the above embodiment of the present invention, and the computer program is controlled and executed by the processor 501. The processor 501 is used to execute the computer program stored in the memory 503 to implement the contents shown in the above method embodiment. Figure 6 The controller 500 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0114] The computer storage medium and controller of the embodiment of the present invention utilize the above-mentioned aerial target tracking method to have the ability to identify and track small targets, realize long-distance target positioning and tracking, and achieve precise strikes in aerial combat.

[0115] It should be noted that the logic and / or steps represented in the flowchart or described in other ways herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection portion with one or more wirings (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways if necessary, and then stored in a computer memory.

[0116] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0117] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0118] In the description of the present invention, it is to be understood that the terms “center”, “longitudinal”, “lateral”, “length”, “width”, “thickness”, “up”, “down”, “front”, “back”, “left”, “right”, “vertical”, “horizontal”, “top”, “bottom”, “inside”, “outside”, “clockwise”, “counterclockwise”, “axial”, “radial”, “circumferential”, etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0119] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0120] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements, unless otherwise clearly defined. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0121] In the present invention, unless otherwise clearly specified and limited, a first feature being "above" or "below" a second feature may mean that the first and second features are in direct contact, or the first and second features are in indirect contact through an intermediate medium. Moreover, a first feature being "above", "above" or "above" a second feature may mean that the first feature is directly above or obliquely above the second feature, or simply means that the first feature is higher in level than the second feature. A first feature being "below", "below" or "below" a second feature may mean that the first feature is directly below or obliquely below the second feature, or simply means that the first feature is lower in level than the second feature.

[0122] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A method for tracking an aerial target, characterized in that: The method comprises: Processing the video frames collected by the guidance head according to the target search range to obtain the target image, wherein the target search range is determined by the initialization target search center or the historical target state information; Performing target detection on the target image to obtain a detected target; Tracking the detection target in the target image according to the target search range to obtain a tracking position of the detection target; Obtaining target state information of the detection target according to the tracking position of the detection target and the historical target state information; The target image includes at least one of a first image, a second image, and a third image, and the video frames collected by the guide head are processed according to the target search range to obtain the target image, including: If the target search range is smaller than the lower limit of the preset size range, taking a screenshot of the video frame with the center of the target search range as the center according to the preset image size to obtain the first image; If the target search range is within a preset size range, compressing the video frame according to a preset image size to obtain the second image; If the target search range is larger than the upper limit of the preset size range, the video frame is edge-filled according to the preset image size and then compressed to obtain the third image; The performing target detection on the target image to obtain the detected target includes: If the target image is the first image, performing feature extraction processing on the target image and then performing target detection to obtain a first target; If the target image is the second image or the third image, performing target detection on the target image using a target detection algorithm to obtain a second target; The first target and / or the second target are fused to obtain the detection target.

2. The aerial target tracking method according to claim 1, characterized in that: The step of performing target detection after performing feature extraction processing on the target image to obtain a first target includes: Using a preset convolution kernel, extracting features of the target image to abstract the target image, wherein the size of the preset convolution kernel is determined by the target scale range; Counting the number of outer circle pixel points whose value of the outer circle pixel of the abstract target in the abstract target image minus the mean value of the inner circle pixels is greater than a preset gradient threshold; The abstract target whose outer circle pixel number is greater than a preset pixel threshold is recorded as the first target.

3. The aerial target tracking method according to claim 1, characterized in that: Tracking the detection target in the target image according to the target search range includes: Tracking and searching the detection target within the target search range, and calculating and obtaining a tracking response value of the detection target; If the tracking response value is greater than a preset response threshold, extracting the center position of the detection target; If the tracking response value is less than a preset response threshold, the target search range is updated according to the detection target and the historical target state information, and the step of tracking and searching the detection target within the target search range is performed.

4. The aerial target tracking method according to claim 1, characterized in that: The target state information includes a target position and a target state, and obtaining the target state information of the detection target according to the tracking position of the detection target and the historical target state information includes: Inferring the position of the detection target according to the historical target position and the historical target state in the historical target state information to determine the inferred position; When the tracking position of the detection target is within a preset range of the inference position, recording the tracking position as the target position of the detection target; The target state of the detection target is calculated based on the tracking position and the historical target position and the historical target state in the historical target state information.

5. The aerial target tracking method according to claim 4, characterized in that: The method further comprises: The target state information is stored, and the target state information is recorded as the historical target state information.

6. An aerial target tracking device, characterized in that: The device comprises: A pre-processing module, used to process the video frames collected by the guidance head according to the target search range to obtain the target image, wherein the target search range is determined by the initialization target search center or the historical target state information; A detection module, used to perform target detection on the target image to obtain a detected target; A tracking module, used to track the detection target in the target image according to the target search range to obtain the tracking position of the detection target; A post-processing module, configured to obtain target state information of the detection target according to the tracking position of the detection target and the historical target state information; The target image includes at least one of a first image, a second image and a third image, and the pre-processing module is used to: if the target search range is smaller than a lower limit of a preset size range, take a screenshot of the video frame with the center of the target search range as the center according to a preset image size to obtain the first image; if the target search range is within a preset size range, compress the video frame according to the preset image size to obtain the second image; if the target search range is larger than an upper limit of the preset size range, fill the edges of the video frame according to the preset image size and then compress it to obtain the third image; The detection module is used to perform target detection after feature extraction processing on the target image to obtain a first target when the target image is the first image; to perform target detection on the target image using a target detection algorithm to obtain a second target when the target image is the second image or the third image; and to fuse the first target and / or the second target to obtain the detected target.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the aerial target tracking method according to any one of claims 1 to 5 is implemented.

8. A controller, comprising a memory and a processor, wherein a computer program is stored in the memory, characterized in that: When the computer program is executed by the processor, the aerial target tracking method according to any one of claims 1 to 5 is implemented.

Images