A target search method based on target prediction
By employing a target prediction sliding window search method in the optoelectronic countermeasures system, combined with a servo mechanism and Kalman filtering, the problem of the tracking and aiming system being unable to quickly lock onto the target due to guidance errors was solved, thus achieving efficient target detection and tracking.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- 中国人民解放军96901部队25分队
- Filing Date
- 2022-10-28
- Publication Date
- 2026-05-26
AI Technical Summary
In existing optoelectronic countermeasures systems, target guidance errors prevent the tracking and aiming system from quickly locking onto the target. Fixed search methods are inefficient and have low information utilization, making it difficult to maintain target tracking during high-speed movement.
A sliding window search method based on target prediction is adopted. Combining prior target information and real-time changes, a sliding window search is performed within a preset range. The servo mechanism and image detection method are used to expand the tracking and aiming search range. Kalman filtering is used to predict the target trajectory and compensate for guidance errors.
At a detection and tracking distance of 20km, it can compensate for guidance errors of hundreds of meters in a short time, improve search efficiency by more than 50%, and achieve fast and efficient target detection and tracking.
Smart Images
Figure CN115825904B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of optoelectronic countermeasures target tracking, and more particularly to a target search method based on target prediction. Background Technology
[0002] With the expansion of optoelectronic countermeasures applications, stable target tracking and aiming based on optoelectronic images has become a crucial function of optoelectronic countermeasures systems. Especially in multi-band laser countermeasures systems based on high-precision pointing, a high-precision and high-efficiency tracking and aiming system is a prerequisite for achieving laser countermeasures. High-precision and high-efficiency target tracking and aiming includes processes such as target guidance, servo rotation, image search, field-of-view switching, and precise tracking. Target guidance primarily relies on long-range detection systems such as optoelectronic search or radar to provide target information, and the remaining steps are completed by a self-closing optoelectronic tracking and aiming system.
[0003] Existing technologies for guiding search and tracking devices in photoelectric detection systems primarily employ coordinate guidance combined with a fixed search method for the tracking system. First, coordinate guidance brings the tracking system near the target's field of view. Then, a fixed search method is used. After the turntable receives target indication information and rotates to the correct position, if no target is found, it waits for new target indication information and continues executing the corresponding target indication information until the target is detected. However, due to factors such as detection errors, the target guidance information may cause the target to be outside the tracking field of view. In such cases, manual adjustment is often used, but manual search is inefficient and cannot guarantee search efficiency and quality.
[0004] like Figure 1-2 As shown, the target is first detected and located using optoelectronic search equipment such as radar or infrared. Based on the relationship between the target's own position and the geometric coordinates of the tracking device, the pitch and azimuth angles that the servo turntable in the tracking device needs to rotate are roughly calculated to guide the optoelectronic center field of view of the tracking device towards the target. Due to the influence of the positioning equipment within the system, the north-finding positioning equipment measuring longitude, and the target's motion characteristics, coordinate methods alone are often insufficient to accurately guide the target into the tracking device's field of view. Therefore, a small-range search method is used to search the adjacent airspace. The principle for selecting the search method is to obtain the largest search range while ensuring no missed scans and meeting the detection probability. Existing search methods often use fixed methods, such as horizontal scanning or spiral scanning, for lateral scanning and retracement. The spiral scanning is similar to scanning the servo turntable in a zigzag pattern. The scanning process is as follows: Assume the initial state is that the system's line of sight points to the left of the heading, and the azimuth is scanned from left to right. While scanning to the right, the pitch angle relative to the horizontal reference continuously increases. After the azimuth scan reaches the far right, the pitch quickly resets to the initial scanning state.
[0005] The main shortcomings of existing technologies are:
[0006] (1) The fixed search pattern of the spiral shape may have too large an overlap rate or inevitably miss scans;
[0007] (2) Due to the randomness of the detection data error of each detector in the photoelectric system and the influence of environmental factors, the fixed search method has a large performance difference under different test conditions, which makes it difficult for operators to make judgments.
[0008] (3) When combined with target information, self-state information, etc., the information utilization rate is low, resulting in low efficiency. Furthermore, there is a disadvantage of losing the target during high-speed target movement and due to small errors in guidance information. Summary of the Invention
[0009] To address the aforementioned technical problems, this invention proposes a target search method based on target prediction, which overcomes the problems of target loss in the tracking system due to guidance deviation and the low efficiency of the fixed search method after guidance.
[0010] According to the above-described scheme of the present invention, this method combines prior target information and real-time changes, employs a sliding window search within a preset range, and combines target motion information to formulate a high-efficiency sliding window mode, thereby expanding the tracking and aiming search range by more than 50%. Furthermore, at a detection and tracking distance of 20km, it can compensate for guidance errors at the hundred-meter level in a short time, while simultaneously solving the problem of low efficiency in traditional tracking and searching. This invention, relying on an optimized electro-optical tracking and aiming system search strategy, achieves efficient searching within a small tracking and aiming field of view, solving the problem of the electro-optical tracking and aiming system failing to find the target due to small errors in guidance information, and improving the OODA loop response time of the electro-optical countermeasures system. This invention addresses the situation where the tracking and aiming system cannot quickly lock onto the target due to guidance deviation. Its basic idea is to use a servo mechanism to guide the electro-optical imaging detector to search using a sliding window method, combined with image detection and other methods, thereby expanding the tracking and aiming search range by more than 50%. At a detection and tracking distance of 20km, it can compensate for guidance errors at the hundred-meter level in a short time. Simultaneously, the efficient sliding window algorithm, combined with target trajectory information, improves search efficiency compared to traditional fixed search methods.
[0011] This invention provides a target search method based on target prediction, the method comprising the following steps:
[0012] Step S1: Detect small, slow targets and predict their positions. The predicted information serves as prior information for the small, slow targets. Based on the prior information, determine the target area where the small, slow targets are located. The target area is the region of the small, slow targets in the image presented by the tracking system. Based on the prior information, adjust the tracking device to reach the external guidance angle of the radar or other system. Subsequently, the video tracking module in the tracking system begins to search for the small, slow targets within the viewport.
[0013] Step S2: If the small, slow target is found in the search view block and can be detected within a preset accuracy, proceed to step S7; otherwise, proceed to step S3.
[0014] Step S3: Based on preset rules, determine a preset region in the target region; divide the preset region into N×N sub-blocks, and select a search starting point based on the prior information, real-time status, and current angle information of the low, small, and slow target, where the search starting point is one of the N×N sub-blocks; use the search starting point as the current search block; proceed to step S4;
[0015] Step S4: Based on the current search area corresponding to the current search block, the video tracking module of the tracking system performs target detection on the imaging image corresponding to the current search area; if a target is detected, proceed to step S7; if no target is detected, proceed to step S5.
[0016] Step S5: If all sub-blocks have been searched, the target cannot be detected. The tracking system is requested to re-guide the process, and the method ends. Otherwise, proceed to step S6.
[0017] Step S6: Based on the detection azimuth, movement speed, and servo turntable status information of the tracking system when detecting the current search block, determine the movement step size and search direction of the servo turntable, thereby enabling the servo turntable to control the tracking system to move into the next sub-block area; the movement step size and search direction are relative to the current state of the servo turntable, the azimuth and pitch step size of the servo turntable, and the movement direction of the servo turntable, respectively; based on the current search block, the movement step size and search direction of the servo turntable, determine the next search sub-block and the next search direction; take the next search sub-block as the current search block and the next search direction as the current search direction, and proceed to step S4;
[0018] Step S7: The low, small, and slow target is detected. The video tracking module outputs the difference value between the target and the servo turntable, so that the tracking system completes the closed-loop tracking of the low, small, and slow target, and the method ends.
[0019] Preferably, in step S3, the preset rule is that in the target area, the area where the photoelectric imaging module is located after coordinate guidance is taken as the central field of view, and the surrounding neighborhood of the central field of view in the target area is obtained, and the surrounding neighborhood is the preset area.
[0020] Preferably, the preset area is divided into N×N sub-blocks, and the area corresponding to each sub-block is the same size as the central field of view imaging area, and there is 50% overlap between the sub-block area and the central field of view, and between the sub-block areas.
[0021] Preferably, the step of selecting the search starting point based on the prior information of the small, slow target, its real-time status, and the current angle information of the servo turntable includes:
[0022] Based on the real-time status of the low, small, and slow target, the search prior error information obtained based on historical data and the prior information of the low, small, and slow target, and the current angle information of the servo turntable, the search starting point is determined.
[0023] Gs=(A1·W1+A2·W2+A3·W3)·G0
[0024] Wherein, Gs is the coordinate of the search starting point, G0 is the current coordinate of the servo turntable, W1, W2, and W3 represent the prior error direction in the prior error information, the current motion direction of the servo turntable, and the current direction of the low, small, and slow target, respectively, and A1, A2, and A3 represent the weights of the three directions, respectively.
[0025] Preferably, the search direction is adjusted according to the directional change gradient of the small, slow target, and the search direction and the movement step size are determined by the following formula:
[0026]
[0027] D = A′·f(v t )
[0028] in, The gradient represents the coordinate change of the small, slow target, and A represents the correction weight. The search direction is represented by D, the movement step size is A′, and f(v) is the correction coefficient. t ) is the target velocity correlation function.
[0029] This invention provides a target search device based on target prediction, the device comprising:
[0030] Initialization module: configured to detect small, slow-moving targets and predict their positions, with the prediction information serving as prior information for the targets. Based on this prior information, the module determines the target area where the targets are located, which is the area of the targets in the image presented by the tracking system. Based on the prior information, the tracking device is adjusted to reach the external guidance angle of the radar or other system. Subsequently, the video tracking module within the tracking system begins to search for the small, slow-moving targets within the viewport.
[0031] First judgment module: configured to trigger the detection module if the small, slow target is found in the search view block and the target can be detected within a preset accuracy; otherwise, trigger the second judgment module.
[0032] The second judgment module is configured to determine a preset region in the target region based on preset rules; divide the preset region into N×N sub-blocks; select a search starting point based on the prior information, real-time status and current angle information of the low, small and slow target, the search starting point is one of the N×N sub-blocks; use the search starting point as the current search block; and trigger the third judgment module.
[0033] The third judgment module is configured to perform target detection on the imaging image corresponding to the current search area based on the current search area corresponding to the current search block; if a target is detected, the detection module is triggered; if no target is detected, the fourth judgment module is triggered.
[0034] The fourth judgment module is configured to, if all sub-blocks have been searched and no target can be detected, request the tracking system to re-guide; otherwise, trigger the movement module.
[0035] The movement module is configured to determine the movement step size and search direction of the servo turntable based on the detection azimuth, movement speed, and status information of the tracking and aiming system when detecting the current search block, thereby enabling the servo turntable to control the tracking and aiming system to move into the next sub-block area; the movement step size and the search direction are relative to the current state of the servo turntable, the azimuth and pitch step size of the servo turntable, and the movement direction of the servo turntable, respectively; based on the current search block, the movement step size and the search direction of the servo turntable, the next search sub-block and the next search direction are determined; the next search sub-block is taken as the current search block, and the next search direction is taken as the current search direction, triggering the third judgment module;
[0036] Detection module: configured to detect the low, small, and slow target, the video tracking module outputs the difference value of the target relative to the servo turntable, so that the tracking system completes closed-loop tracking of the low, small, and slow target.
[0037] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, the preferred embodiments of the present invention are described in detail below with reference to the accompanying drawings. Attached Figure Description
[0038] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention, and the invention is illustrated by the following drawings. In the drawings:
[0039] Figure 1 This is a schematic diagram of the existing radar / infrared search and guidance tracking servo turntable process;
[0040] Figure 2 This is a schematic diagram of a spiral scanning servo turntable in existing technology.
[0041] Figure 3 This is a flowchart illustrating the implementation process of the tracking and aiming system of the present invention.
[0042] Figure 4 This is a schematic diagram illustrating the applicable scenarios of the present invention;
[0043] Figure 5 A flowchart of a target search method based on target prediction within a preset range, representing one embodiment of the present invention;
[0044] Figure 6 This is a schematic diagram illustrating the direction selection of a target search method based on target prediction according to one embodiment of the present invention. Detailed Implementation
[0045] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.
[0046] definition:
[0047] Detection (sub)system: generally refers to long-range detection equipment formed by a single device or combination of devices such as radar, infrared detection equipment, and visible light detection equipment;
[0048] Tracking and aiming system: generally refers to a weapon tracking and aiming system composed of a servo turntable, an optoelectronic imaging module, a video tracking module, etc., which guides the laser to continuously aim at the target in an optoelectronic countermeasures system;
[0049] Optoelectronic imaging module: refers to an optoelectronic imaging device composed of infrared detectors and visible light detectors;
[0050] Video tracking module: also known as a video tracker, it can perform target detection on the imaging video of the photoelectric imaging module and output the difference value to control the servo turntable to track the inspected target.
[0051] First, common low-altitude, slow-moving, small unmanned aerial vehicles (UAVs) are used as targets for the electro-optical system to detect targets, combined with... Figure 3-4The following is a flowchart illustrating a target search method based on target prediction, according to one embodiment of the present invention. The method includes the following steps:
[0052] Step S1: Detect small, slow-moving targets and predict their positions. The predicted information serves as prior information for the targets. Based on the prior information, determine the target area where the targets are located. The target area is the region of the targets in the image presented by the tracking system. Based on the prior information, adjust the tracking device to reach the external guidance angle of the radar or other system. Then, the video tracking module in the tracking system begins to search for the small, slow-moving targets within the viewport.
[0053] In this embodiment, the external guidance angle is achieved by first detecting the low, small, and slow target using an infrared search device or a radar detection device, and then guiding the tracking system to align with the low, small, and slow target based on coordinates and other information. Subsequently, the video tracker in the tracking system begins to search for the low, small, and slow target within the view block.
[0054] Step S2: If the small, slow target is found in the search view block and can be detected within a preset accuracy, proceed to step S7; otherwise, proceed to step S3.
[0055] Step S3: Based on preset rules, determine a preset region in the target region; divide the preset region into N×N sub-blocks, and select a search starting point based on the prior information, real-time status, and current angle information of the low, small, and slow target, where the search starting point is one of the N×N sub-blocks; use the search starting point as the current search block; proceed to step S4;
[0056] Step S4: Based on the current search area corresponding to the current search block, the video tracking module of the tracking system performs target detection on the imaging image corresponding to the current search area; if a target is detected, proceed to step S7; if no target is detected, proceed to step S5.
[0057] Step S5: If all sub-blocks have been searched, the target cannot be detected. The tracking system is requested to re-guide the process, and the method ends. Otherwise, proceed to step S6.
[0058] Step S6: Based on the detection azimuth, movement speed, and servo turntable status information of the tracking system when detecting the current search block, determine the movement step size and search direction of the servo turntable, thereby enabling the servo turntable to control the tracking system to move into the next sub-block area; the movement step size and search direction are relative to the current state of the servo turntable, the azimuth and pitch step size of the servo turntable, and the movement direction of the servo turntable, respectively; based on the current search block, the movement step size and search direction of the servo turntable, determine the next search sub-block and the next search direction; take the next search sub-block as the current search block and the next search direction as the current search direction, and proceed to step S4;
[0059] In this embodiment, the movement step size refers to the azimuth and pitch step size of each search movement of the servo turntable, corresponding to changes in the target imaging range. The search direction is relative to the current state of the servo turntable, and the movement direction of the servo turntable corresponds to the direction of change in the search airspace.
[0060] Step S7: The low, small, and slow target is detected. The video tracking module outputs the difference value between the target and the servo turntable, so that the tracking system completes the closed-loop tracking of the low, small, and slow target, and the method ends.
[0061] like Figure 5 As shown, this method is applicable to an optoelectronic detection system consisting of a detection system and a tracking system. The target position information obtained from the detection system guides the tracking system to determine a preset area. A search is then conducted within this preset area. During the search, Kalman filtering is introduced to predict target motion, enabling the tracking device to accurately detect and track the target.
[0062] In step S1, the detection system, including radar / infrared devices, detects low-lying, small, and slow-moving targets and provides their geographical location. Subsequently, based on the coordinate relationship between the detection system and the tracking system, a coordinate matrix transformation is performed on the target's position to obtain information such as the target's azimuth, elevation, distance, and velocity relative to the tracking system. The sensor automatic control system within the tracking device's servo turntable controls the servo turntable to align with the target's location, guiding the photoelectric imaging module and video tracking module to image and detect the target's area. However, due to factors such as guidance errors and servo rotation errors, there will be a certain field-of-view deviation between the image area formed by the tracking system after coordinate guidance and the actual target area.
[0063] In step S2, if a target is detected, that is, if the low, small, and slow target is detected in the field of view of the photoelectric imaging module of the tracking device, the low, small, and slow target is displayed in the form of an infrared / visible light image. The video tracking module can then detect the low, small, and slow target within the current detection field of view and achieve closed-loop tracking.
[0064] Step S3, addressing the situation where coordinate guidance cannot directly locate the target, is also one of the core technical points of this invention. The preset rule is that, within the target area, the area where the photoelectric imaging module is located after coordinate guidance is taken as the central field of view, and the surrounding neighborhood of the central field of view within the target area is obtained; the surrounding neighborhood is the preset area.
[0065] Furthermore, the preset area is divided into N×N sub-blocks, and the area corresponding to each sub-block is the same size as the central field of view imaging area, and there is a 50% overlap between the sub-block area and the central field of view, and between the sub-block areas themselves.
[0066] Then, the optimal starting search point is selected from the N×N sub-block regions. Its basic core is to predict the target's movement direction based on the detected target information, and by selecting the optimal starting search point, the tracking system can detect the target with the highest probability when both the target and the tracking system are in motion.
[0067] Furthermore, the selection of the search starting point based on the prior information of the low, small, and slow target, its real-time status, and the current angle information of the servo turntable includes:
[0068] Based on the real-time status of the low, small, and slow target, the search prior error information obtained based on historical data and the prior information of the low, small, and slow target, and the current angle information of the servo turntable, the search starting point is determined.
[0069] Gs=(A1·W1+A2·W2+A3·W3)·G0
[0070] Wherein, Gs represents the coordinates of the search starting point, G0 represents the current coordinates of the servo turntable, W1, W2, and W3 represent the direction of the prior error in the search prior error information, the current direction of the servo turntable's movement, and the current direction of the low, small, and slow target, respectively. A1, A2, and A3 represent the weights of the three directions, respectively. These three factors are combined to calculate the optimal search starting point, and the search starting point is determined based on its coordinates. The coordinates of the optimal starting point are located within the search starting point.
[0071] In step S4, the video tracking module performs target detection in each sub-block region. The basic idea is to use image detection algorithms to extract the target from the background image. Since the most commonly used target detection algorithm in practical applications is the moving target detection algorithm, taking frame difference as an example, each sub-block needs at least 3 consecutive frames. If the video frame rate is 25 frames / s, then each region needs to remain in the image for 0.12 seconds. If a current frame detection algorithm (such as the YOLO series of artificial intelligence algorithms) is used, only 1 frame is needed to complete the detection.
[0072] Step S6 indicates that no target is detected in the current field of view of the tracking system. The servo turntable needs to move to drive the tracking system to detect the next sub-block. Therefore, it is necessary to determine the direction and compensation of the servo turntable movement, which is represented by the image sliding window step size and direction. Traditional search sliding windows traverse around the central field of view in a certain order until the target is found. The method of this invention combines a small-range search near the center of the field of view with a tracking prediction rotation based on Kalman filtering during the search process. It introduces target information, predicts the target trajectory, and achieves fast and efficient search within a preset range in the area. It pulls the target to the center of the image field of view for display, continuous tracking and identification. This enables the photoelectric system to automatically detect, track, monitor and identify aerial targets under various weather conditions, achieving fully automated system operation. At the same time, it can correct situations where the target is not in the tracking field of view due to small guidance errors and other factors, improving target search efficiency and increasing the target detection probability. In this invention, the movement step size and search direction of the servo turntable are determined based on the detection orientation, movement speed, and status information of the servo turntable when the tracking and aiming system detects the current search block, combined with the tracking and prediction method of Kalman filtering.
[0073] In this embodiment, the sliding window search starts from the corner of the field of view. Since the infrared thermal imager lacks a backscan component, it cannot compensate for image drift caused by the turntable movement, affecting target detection. Therefore, a step-by-step turntable rotation strategy is adopted. By adjusting the servo turntable position loop control, a step-by-step sliding window search is achieved.
[0074] like Figure 6 As shown, in step S6:
[0075] The movement step size is relative to the current state of the servo turntable. The azimuth and pitch step sizes of the servo turntable movement are relative to the azimuth and pitch motion of the currently searched sub-block, i.e., the motion of the sliding window. Target estimation is performed based on Kalman filtering, including inferring the trajectory of a low-speed, small target based on the target's motion direction, speed, and prior error information. Simultaneously, trajectory matching is performed according to the sliding window direction, enabling the tracking and aiming system to quickly and efficiently search for the target's sub-region during target movement.
[0076] The search direction is the servo turntable movement direction relative to the search direction of the current search block. In this embodiment, the search direction is adjusted according to the gradient of the target direction change.
[0077] Furthermore, the search direction is adjusted according to the gradient of the target direction change, and a certain overlap rate (i.e., rotation step size) is maintained in each sliding window to avoid loss due to target movement. The target direction is determined by the following formula:
[0078]
[0079] D = A′·f(v t )
[0080] in, The gradient representing the coordinate change of the small, slow target is obtained from the system's infrared search / radar detectors, and A represents the correction weight. The target direction is represented by D, the movement step size is A′, and f(v) is the correction coefficient. t ) is the target velocity correlation function.
[0081] The preset region is divided into N×N blocks, and the tracking field of view is searched in each block. Based on the information from the previous block and the continuously returned detection information, the best matching block and the starting point for the next search are found. Furthermore, the maximum step size of the search area can be set to reduce delays in system operation due to the inability to acquire the target. After each search, the image of each field of view block is inspected.
[0082] In this embodiment, due to the uncertainty of the target's location information and the randomness of the search results within the preset range, it is necessary to consider the real-time update of the target's location information and adopt a more reasonable search decision method and servo rotation angle parameters to ensure that the tracking and aiming device reduces the uncertainty of the task area in the shortest possible time and discovers the target as quickly as possible. Therefore, Kalman filtering is introduced to estimate the target's motion, that is, by predicting the target's motion trajectory, the selection of the next motion state of the servo turntable is driven, guiding the servo turntable to rotate synchronously. The tracking prediction rotation based on Kalman filtering, that is, during the sliding window search, the system predicts the target's motion vector through Kalman filtering, guides the servo turntable to rotate synchronously, and compensates for the search offset caused by the target's motion during the sliding window time. In the above formula, this is represented by the correction values A and A′.
[0083] Due to the diverse types and significant maneuverability and angular velocity of small, slow-moving aerial targets, ground-based search and tracking systems require high levels of autonomous target identification and tracking accuracy. This invention employs Kalman filtering to estimate and filter target velocity and acceleration information, predicting the servo rotation angle that forms the target trajectory. This angle serves as the input for servo feedforward compensation, eliminating errors caused by target velocity and acceleration and improving estimation accuracy.
[0084] Furthermore, during the search within the preset area, the image processor within the tracking system can process the infrared images acquired by the photoelectric turret sensor and automatically detect and track the target. After acquiring real-time visible light / infrared images, the image processor first preprocesses the image using a target contrast enhancement algorithm and suppresses the target background. Then, it segments the image and performs false alarm suppression, correlates multiple frames of the sequence images to extract the target, and detects and locks the target with the highest signal-to-noise ratio. This algorithm can run on embedded devices, enabling fast and accurate target detection and guiding the tracking system to complete closed-loop tracking.
[0085] Step S7, detecting the small, slow target, includes:
[0086] After detecting the small, slow-moving target, the video tracking module simultaneously begins to detect targets within the field of view. Once the target is detected, the servo turntable is controlled to lock onto the target.
[0087] The following examples illustrate the target search method based on target prediction of the present invention.
[0088] The tracking and aiming system exemplified in this invention consists of hardware such as a high-precision servo turntable, an infrared thermal imager, and a video tracker. The infrared thermal imager typically has a large field of view (3° × 2.4°) and a small field of view (1.5 × 1.2°). Based on this hardware, a small-area search function is designed for the high-precision tracking and aiming system, enabling target search within a 5° × 5° area.
[0089] When the tracking system receives a search area command, such as using a large infrared field of view (3°×2.4°), in order to complete a 5°×5° search area, the search starts from the left corner of the area based on factors such as the target's prior information, real-time status, prior search error information, and turntable rotation. Taking into account the turntable's azimuth / elevation tracking angular velocity and infrared detection frequency, assuming the turntable rotates at a speed of 10 degrees / second, after the azimuth search is completed, the elevation baseline is adjusted, while ensuring a 15% overlap rate in the elevation direction. Therefore, within the entire search area, the turntable needs to change direction twice and adjust the elevation baseline twice. The direction change and elevation baseline adjustment are completed synchronously, with each adjustment taking 0.3 seconds. Therefore, the entire search time takes 2.1 seconds to complete a 5°×6° area scan. During the search, the target is detected in real time, and if no target is found, the system returns to its original position.
[0090] If a small infrared field of view (1.5°×1.2°) is used, to complete a 5°×5° search area, the search begins from the left corner of the area, with the turntable rotating at a speed of 10 degrees / second. After the azimuth search is completed, the elevation baseline is adjusted to ensure a 10% overlap in the elevation area. Therefore, the turntable needs to perform three rotations and three elevation baseline adjustments within the entire search area. The rotation and elevation baseline adjustments are performed synchronously, with each adjustment taking 0.2 seconds. Thus, the entire search time is 2.6 seconds to complete a 5°×5.3° area scan. During the search, the target is detected in real time, and if no target is found, the turntable returns to its original position. Based on the turntable's rotation acceleration, a single azimuth and elevation rotation can be completed in 0.3 seconds. Considering the overlap rate of the azimuth and elevation fields of view and the infrared staring detection time, a 5°×6° area search can be completed in 1.8 seconds under a large infrared field of view, and in 2.8 seconds under a small infrared field of view.
[0091] This invention provides a target search device based on target prediction, the device comprising:
[0092] Initialization module: configured to detect small, slow-moving targets and predict their positions, with the prediction information serving as prior information for the targets. Based on this prior information, the module determines the target area where the targets are located, which is the area of the targets in the image presented by the tracking system. Based on the prior information, the tracking device is adjusted to reach the external guidance angle of the radar or other system. Subsequently, the video tracking module within the tracking system begins to search for the small, slow-moving targets within the viewport.
[0093] First judgment module: configured to trigger the detection module if the small, slow target is found in the search view block and the target can be detected within a preset accuracy; otherwise, trigger the second judgment module.
[0094] The second judgment module is configured to determine a preset region in the target region based on preset rules; divide the preset region into N×N sub-blocks; select a search starting point based on the prior information, real-time status and current angle information of the low, small and slow target, the search starting point is one of the N×N sub-blocks; use the search starting point as the current search block; and trigger the third judgment module.
[0095] The third judgment module is configured to perform target detection on the imaging image corresponding to the current search area based on the current search area corresponding to the current search block; if a target is detected, the detection module is triggered; if no target is detected, the fourth judgment module is triggered.
[0096] The fourth judgment module is configured to, if all sub-blocks have been searched and no target can be detected, request the tracking system to re-guide; otherwise, trigger the movement module.
[0097] The movement module is configured to determine the movement step size and search direction of the servo turntable based on the detection azimuth, movement speed, and status information of the tracking and aiming system when detecting the current search block, thereby enabling the servo turntable to control the tracking and aiming system to move into the next sub-block area; the movement step size and the search direction are relative to the current state of the servo turntable, the azimuth and pitch step size of the servo turntable, and the movement direction of the servo turntable, respectively; based on the current search block, the movement step size and the search direction of the servo turntable, the next search sub-block and the next search direction are determined; the next search sub-block is taken as the current search block, and the next search direction is taken as the current search direction, triggering the third judgment module;
[0098] Detection module: configured to detect the low, small, and slow target, the video tracking module outputs the difference value of the target relative to the servo turntable, so that the tracking system completes closed-loop tracking of the low, small, and slow target.
[0099] This invention provides a target search system based on target prediction, the main components of which include:
[0100] Detection (sub)system: generally refers to long-range detection equipment formed by a single device or combination of devices such as radar, infrared detection equipment, and visible light detection equipment;
[0101] Tracking and aiming system: generally refers to a weapon tracking and aiming system composed of a servo turntable, an optoelectronic imaging module, a video tracking module, etc., which guides the laser to continuously aim at the target in an optoelectronic countermeasures system;
[0102] Optoelectronic imaging module: refers to an optoelectronic imaging device composed of infrared detectors and visible light detectors;
[0103] Video tracking module: also known as a video tracker, it can perform target detection on the imaging video of the photoelectric imaging module and output the difference value to control the servo turntable to track the inspected target.
[0104] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention shall still fall within the scope of the technical solution of the present invention.
Claims
1. A target search method based on target prediction, characterized in that, The method includes the following steps: Step S1: Detect small, slow targets and predict their positions. The predicted information serves as prior information for the small, slow targets. Based on the prior information, determine the target area where the small, slow targets are located. The target area is the region of the small, slow targets in the image presented by the tracking system. Based on the prior information, adjust the tracking device to reach the external guidance angle of the radar or other system. Subsequently, the video tracking module within the tracking system begins to search for the small, slow targets within the viewport. Step S2: If the small, slow target is found in the search view block and can be detected within a preset accuracy, proceed to step S7; otherwise, proceed to step S3. Step S3: Based on preset rules, determine a preset region in the target region; divide the preset region into N×N sub-blocks, and select a search starting point based on the prior information, real-time status, and current angle information of the low, small, and slow target, where the search starting point is one of the N×N sub-blocks; use the search starting point as the current search block; proceed to step S4; Step S4: Based on the current search area corresponding to the current search block, the video tracking module of the tracking system performs target detection on the imaging image corresponding to the current search area; if a target is detected, proceed to step S7; if no target is detected, proceed to step S5. Step S5: If all sub-blocks have been searched, the target cannot be detected. The tracking system is requested to re-guide the process, and the method ends. Otherwise, proceed to step S6. Step S6: Based on the detection azimuth, movement speed, and servo turntable status information of the tracking system when detecting the current search block, determine the movement step size and search direction of the servo turntable, thereby enabling the servo turntable to control the tracking system to move into the next sub-block area; the movement step size and search direction are relative to the current state of the servo turntable, the azimuth and pitch step size of the servo turntable, and the movement direction of the servo turntable, respectively; based on the current search block, the movement step size and search direction of the servo turntable, determine the next search sub-block and the next search direction; take the next search sub-block as the current search block and the next search direction as the current search direction, and proceed to step S4; Step S7: Upon detecting the small, slow target, the video tracking module outputs the difference between the target and the servo turntable, enabling the tracking system to complete closed-loop tracking of the small, slow target, and the method ends. In step S3, the preset rule is that in the target area, the area where the photoelectric imaging module is located after coordinate guidance is taken as the central field of view, and the surrounding neighborhood of the central field of view in the target area is obtained. The surrounding neighborhood is the preset area. The preset area is divided into N×N sub-blocks. The preset area corresponding to each sub-block is the same size as the central field of view imaging area, and there is a 50% overlap between the preset area corresponding to the sub-block and the central field of view, and between the preset areas corresponding to the sub-blocks. The search direction is adjusted according to the gradient of the direction change of the small, slow target, and the search direction and the movement step size are determined by the following formula: in, The gradient represents the coordinate change of the small, slow target, and A represents the correction weight. This indicates the search direction, and D is the movement step size. For correction factor, This is the target velocity correlation function.
2. The method as described in claim 1, characterized in that, The selection of the search starting point based on the prior information, real-time status, and current angle information of the low, small, and slow target includes: Based on the real-time status of the low, small, and slow target, the search prior error information obtained based on historical data and the prior information of the low, small, and slow target, and the current angle information of the servo turntable, the search starting point is determined. Gs = (A1·W1+ A2·W2+ A3·W3)·G0 Wherein, Gs is the coordinate of the search starting point, G0 is the current coordinate of the servo turntable, W1, W2, and W3 represent the prior error direction in the prior error information, the current motion direction of the servo turntable, and the current direction of the low, small, and slow target, respectively, and A1, A2, and A3 represent the weights of the three directions, respectively.
3. A target search apparatus based on target prediction, used to execute the method of any one of claims 1-2, characterized in that, The device includes: Initialization module: configured to detect small, slow-moving targets and predict their positions, with the prediction information serving as prior information for the targets. Based on this prior information, the module determines the target area where the targets are located, which is the area of the targets in the image presented by the tracking system. Based on the prior information, the tracking device is adjusted to reach the external guidance angle of the radar or other system. Subsequently, the video tracking module within the tracking system begins to search for the small, slow-moving targets within the viewport. First judgment module: configured to trigger the detection module if the small, slow target is found in the search view block and the target can be detected within a preset accuracy; otherwise, trigger the second judgment module. The second judgment module is configured to determine a preset region in the target region based on preset rules; divide the preset region into N×N sub-blocks; select a search starting point based on the prior information, real-time status and current angle information of the low, small and slow target, the search starting point is one of the N×N sub-blocks; use the search starting point as the current search block; and trigger the third judgment module. The third judgment module is configured to perform target detection on the imaging image corresponding to the current search area based on the current search area corresponding to the current search block; if a target is detected, the detection module is triggered; if no target is detected, the fourth judgment module is triggered. The fourth judgment module is configured to, if all sub-blocks have been searched and no target can be detected, request the tracking system to re-guide; otherwise, trigger the movement module. The movement module is configured to determine the movement step size and search direction of the servo turntable based on the detection azimuth, movement speed, and status information of the tracking and aiming system when detecting the current search block, thereby enabling the servo turntable to control the tracking and aiming system to move into the next sub-block area; the movement step size and the search direction are relative to the current state of the servo turntable, the azimuth and pitch step size of the servo turntable, and the movement direction of the servo turntable, respectively; based on the current search block, the movement step size and the search direction of the servo turntable, the next search sub-block and the next search direction are determined; the next search sub-block is taken as the current search block, and the next search direction is taken as the current search direction, triggering the third judgment module; Detection module: configured to detect the low, small, and slow target, the video tracking module outputs the difference value of the target relative to the servo turntable, so that the tracking system completes closed-loop tracking of the low, small, and slow target.