Intelligent target searching system and target searching method based on deep learning

Abstract

The invention discloses an intelligent target searching system and method based on deep learning, relates to the technical field of artificial intelligence, computer vision and automatic target scoring, solves the problems that a traditional target scoring system is insufficient in target searching precision, prone to being influenced by the environment, dependent on manual adjustment and the like, and is particularly suitable for complex scenes with multi-target-position and multi-camera layout. The method is mainly applied to shooting training scenes such as armies and police teams, the automatic and intelligent target searching process is achieved through the deep learning technology, and the positioning precision, the real-time performance and the environment adaptability of the system can be improved. According to the intelligent target searching method, intelligent target searching of a multi-target-position multi-camera system is realized through multi-module cooperation and integration of data acquisition, deep learning detection, rotary table accurate control and feedback adjustment. According to the method, the target searching speed and efficiency are remarkably improved on the premise that the precision requirement is met, and the method has wide practical value and good application prospects.

Description

Technical Field

[0001] The present invention relates to the fields of artificial intelligence, computer vision and automatic target reporting technology, and in particular to an intelligent target finding system and a target finding method based on deep learning. Background Art

[0002] With the development of optoelectronic technology and intelligent recognition algorithms, non-contact automatic target-marking systems based on vision are gradually replacing contact target-marking systems with high usage costs. In the automatic target-marking system based on vision, the layout of multiple targets and multiple cameras is generally "one-to-one correspondence". In the system of multiple targets and camera equipment, it is crucial for the task of the automatic target-marking system to achieve accurate alignment and automatic target search of each camera to a specific target. This process is usually called "target search", which refers to the process in which each camera finds and aligns the predetermined target in a complex scene so that the target is in the center or optimal position in its field of view. The accuracy and efficiency of target search directly affect the performance of the entire target-marking system. Traditional target search technology mostly relies on mechanical adjustment, manual calibration or simple image processing algorithms, such as locating the target by features such as color, shape, and edge detection. These methods usually require manual adjustment and pre-calibration, and are greatly affected by ambient lighting, target posture and occlusion factors, making it difficult to achieve automatic alignment in dynamic or complex environments. In addition, traditional methods often lack effective feature distinction means in multi-target scenes, and are prone to false detection or misalignment problems, which in turn affects the stability and accuracy of the system. With the rapid development of deep learning technology, target detection methods based on convolutional neural networks (CNNs) have brought new possibilities to target-finding tasks. Deep learning technology can automatically learn the multi-level features of images, allowing cameras to effectively identify and locate specific targets under different lighting and complex backgrounds. How to design an efficient and accurate deep learning target-finding method so that each camera can accurately find the corresponding target in real-time dynamic conditions and maximize the target area within its field of view has become a hot topic and difficulty in current research.

[0003] To this end, there is an urgent need for an efficient and accurate intelligent target search method based on deep learning, which can improve the real-time performance of the model while ensuring recognition accuracy, support intelligent automatic alignment of multi-camera systems, and adapt to diverse target layouts.

[0004] The intelligent target-finding system and target-finding method of the present invention are used for target reporting equipment with multiple target positions and multiple cameras for shooting training in the military, police force, and military units. The method receives real-time frame images from multiple cameras, extracts semantic information of the images using multiple convolution operations, identifies the targets, and outputs the target miss distance information to the main control system of the equipment. The main control system then drives the turntable to rotate according to the miss distance information, and repeats this process until all target positions appear in the center of the field of view of the target camera and fill the entire field of view to the maximum extent. Summary of the invention

[0005] The present invention provides an intelligent target-finding system and method based on deep learning, which aims to solve the problems of insufficient target-finding accuracy, susceptibility to environmental influences, and reliance on manual adjustment in traditional target-finding systems, and is particularly suitable for complex scenes with multiple target positions and multiple cameras. The method is mainly used in shooting training scenes such as the military and police forces. By adopting deep learning technology to realize an automated and intelligent target-finding process, the system's positioning accuracy, real-time performance, and environmental adaptability can be improved.

[0006] An intelligent target-seeking system based on deep learning, the system includes a data acquisition module, a feature extraction module, a target detection module, a deviation calculation module, a turntable adjustment module and a feedback control module;

[0007] The data acquisition module is used to obtain the real-time video frame image of each camera and transmit it to the feature extraction module;

[0008] The feature extraction module extracts features from the video frame image, obtains edge, texture and shape features of the target image, and obtains a target object feature map;

[0009] The target detection module uses a target detection algorithm based on deep learning to identify the target target feature map, obtains the coordinate position of the target target in the feature map, that is, the target center position; and transmits the target center position to the deviation calculation module;

[0010] The deviation calculation module calculates the deviation value of the target center coordinate relative to the camera field of view center, i.e., the miss distance, according to the target center coordinate detected by the target detection module;

[0011] The turntable adjustment module drives the turntable where the camera is located to adjust the direction and angle in real time according to the miss distance information obtained by the deviation calculation module, so as to ensure that the target position is always in the best field of view position of the camera;

[0012] The feedback control module is used to continuously adjust and optimize the control parameters of the turntable according to the position information of the target in the current field of view during the process of the camera aligning the target, until the target is located in the central field of view of the camera and maximizes the coverage of the entire field of view area; when the camera completes the accurate alignment of the target, the feedback control module will output a completion signal.

[0013] The present invention also provides an intelligent target-seeking method based on deep learning, which is implemented by the following steps:

[0014] Step 1: Obtain high-resolution video frame images transmitted in real time by the camera through the data acquisition module;

[0015] Step 2: Preprocess the high-resolution video frame image obtained in step 1 to obtain a low-resolution video frame image, and convert the RGB image into a grayscale image;

[0016] Step 3: Use the CNN network model as a feature extraction module to perform feature extraction on the low-resolution video frame image, extract the edge, texture and shape features of the target image, and obtain a target feature map;

[0017] Step 4: Use the target detection module to identify the target position;

[0018] Detecting the coordinate position of the target object in the target object feature map, after completing the target identification, transmitting the target center coordinate to the deviation calculation module, and the deviation calculation module calculates the offset between the camera field of view center and the target center coordinate according to the target center coordinate, that is, the miss distance;

[0019] Step 5, the deviation calculation module transmits the miss distance information to the turntable adjustment module, and the turntable adjustment module drives the turntable where the camera is located to adjust the direction and angle in real time according to the miss distance information to ensure that the target position is always in the best field of view position of the camera;

[0020] When the camera is aligned with the target, the feedback control module continuously adjusts and optimizes the control parameters of the turntable according to the position information of the target in the current field of view, until the target is located in the center field of view of the camera and covers the entire field of view area to the maximum extent; when the camera completes the accurate alignment of the target, the feedback control module will output a completion signal.

[0021] Beneficial effects of the present invention:

[0022] 1. Intelligent automatic alignment. The intelligent target search method described in the present invention can realize automatic alignment in complex scenes, replacing traditional manual calibration and mechanical adjustment. Through the multi-level feature extraction capability of CNN, the system can effectively identify the target position under different lighting, angles and occlusion conditions, significantly improving the target search accuracy.

[0023] 2. Real-time feedback and precise control. By introducing the deviation calculation and feedback control module, the present invention can adjust the camera field of view in real time so that the target position is always in the center. The PID control algorithm ensures that the turntable responds smoothly during the adjustment process, reduces jitter and unnecessary waiting time, and ensures the efficiency and real-time performance of the system.

[0024] 3. Application of lightweight detection model. The present invention adopts lightweight target detection models, such as YOLO or MobileNet-SSD, to ensure efficient detection speed, making the system suitable for embedded platforms or low-power edge computing devices. In addition, the lightweight design of the model reduces the system computing overhead and supports real-time target search at high frame rates.

[0025] Fourth, modular design and strong applicability. The intelligent target-seeking method of the present invention can be flexibly applied to different types of multi-target and multi-camera layouts through modular design. Each module operates independently and cooperates with each other. Through the complete process of data acquisition, target detection, deviation calculation and feedback control, it can be promoted and applied in military shooting training, security monitoring, automatic target reporting system and other scenarios.

[0026] 5. High adaptability and robustness. The present invention not only improves the system's target-finding accuracy but also enhances its robustness in complex environments through the combination of CNN and PID feedback control. Even in the case of ambient lighting, target occlusion or field of view changes, the system can still effectively achieve automatic target alignment and real-time tracking.

[0027] In summary, the intelligent target search method described in the present invention realizes intelligent target search for a multi-target multi-camera system through multi-module collaboration, integrated data acquisition, deep learning detection, turntable precision control and feedback adjustment. This method significantly improves the target search speed and efficiency while meeting the accuracy requirements, and has broad practical value and good application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 This is a schematic diagram of an intelligent target-seeking system based on deep learning according to the present invention;

[0029] Figure 2 This is a schematic diagram of the layout of an intelligent target-seeking system based on deep learning according to the present invention;

[0030] Figure 3 This is the schematic diagram of the network structure of the feature extraction module and the target detection module;

[0031] Figure 4 This is the schematic diagram of the deviation calculation module;

[0032] Figure 5 is a flow chart of the feedback control module;

[0033] Figure 6 This is a flow chart of an intelligent target-seeking method based on deep learning according to the present invention;

[0034] Figure 7 This is a schematic diagram of target position recognition under different environmental conditions using a deep learning-based intelligent target-seeking system described in the present invention. DETAILED DESCRIPTION

[0035] Specific implementation method 1. Combination Figures 1 to 5The present embodiment is described as an intelligent target-seeking system based on deep learning. The intelligent target-seeking system includes an automatic target-seeking device with multiple target positions, multiple cameras (data acquisition modules) and a turntable assembly. The spatial layout between the automatic target-seeking device and the multiple target positions is as follows: Figure 2 As shown, the humanoid target and the digital representation are arranged in an upper and lower manner. The longitudinal distance between the automatic target-seeking device and the target supports 10-15m, the interval distance between multiple targets supports 2-4m, and supports the expansion of the number of custom targets, showing the overall layout of the multi-camera, multi-target automated target-seeking system. This embodiment uses deep learning and feedback control to achieve high-precision, high-real-time automatic alignment of targets. The automatic target-seeking device communicates with the industrial computer through a switch; the industrial computer is integrated with a feature extraction module, a target detection module, a deviation calculation module, and a turntable control module (turntable adjustment module and feedback control module);

[0036] The data acquisition module is used to obtain real-time video frame images of each camera, and transmit the real-time image frames to the industrial computer system according to the set frame rate. The collected image frames will directly enter the feature extraction module.

[0037] In this embodiment, the camera is mounted on a turntable and can freely adjust the direction and angle to adapt to the position of different targets. In order to improve the stability and adaptability of data acquisition, in this embodiment, a highly sensitive CMOS or CCD sensor camera is preferably used to ensure that the target image can be clearly captured in complex environments such as low light and strong light to ensure image quality. The collected video frame image will be directly transmitted to the subsequent module to ensure that the system has high real-time performance. The data acquisition module sets a certain frame rate to ensure that the collected video frame has a sufficient refresh rate to facilitate subsequent feature extraction and target detection.

[0038] The feature extraction module uses a convolutional neural network (CNN) to perform multi-level feature extraction on the video frame images captured by the camera. First, the video frame images are processed by multiple layers of convolution and activation functions to extract low-level edge and contour features; then the feature dimension is reduced through the pooling layer and the computational efficiency of the model is increased. Figure 3 As shown in the figure, the feature extraction module contains multiple convolutional layers and pooling layers, which can extract the edge, texture and shape features of the target image. The baseline model used in the target-seeking system is Yolov10. On its basis, customized improvements and modifications are made according to the use environment of the equipment and the characteristics of the target, including the WTB module and the CorrdAtt module. The WTB module is a convolution operation based on wavelet transform. It can extract the features of the target more finely and improve the detection accuracy without increasing the complexity of the model. For detailed introduction of other modules, such as Conv module and C2f module, please refer to the original text of Yolov10 author (Link: https: / / arxiv.org / abs / 2405.14458 ), which will not be elaborated here. This feature extraction module can adapt to different lighting, angles and occlusion conditions to ensure accurate identification of the target even in complex backgrounds. Compared with traditional manual feature extraction methods, the CNN feature extraction method used in this module has stronger adaptability and generalization ability, and can accurately identify the target in a variety of environments, greatly improving the robustness and accuracy of the system.

[0039] The target detection module identifies the target object based on the deep learning target detection algorithm and obtains the coordinate position of the target in the image. The target detection module can accurately predict the bounding box position of the target object in the image by combining the bounding box regression algorithm and the loss function convergence. After completing the target position identification, the detection result (x, y, K) is transmitted to the deviation calculation module.

[0040] like Figure 3 As shown, in this embodiment, the target detection module adopts a lightweight deep learning model, such as YOLO or MobileNet-SSD, to ensure the balance between detection accuracy and real-time performance of the system. The module uses the feature map extracted by CNN to predict the coordinate position of the target through bounding box regression. Specifically, it includes preprocessing of the input image, forward propagation of the model, bounding box generation and coordinate extraction. During the forward propagation process, the target detection module filters the background noise according to the set threshold and only extracts the target area with high confidence. The detected target center coordinates and bounding box coordinates will be transmitted to the deviation calculation module for calculating the offset of the target relative to the center of the camera field of view.

[0041] like Figure 4 As shown, Figure 4 The method for expressing the position of the target in the camera field of view is to use the camera field of view as the two-dimensional coordinate origin (0,0), the X-axis range (-0.5,0.5), and the Y-axis range (-0.5,0.5). The position of the target center relative to the camera field of view center is represented by a two-dimensional coordinate system, where the detection output result is (x, y, K). In the figure, the coordinates of point A are (x, y), and K = S 靶标 / S 视场 , the coordinates of point A represent the position of the target center in the coordinate system established in the image as shown in the figure, and x and y represent the horizontal and vertical deviation values ​​respectively; the target-seeking task requires the target to occupy 8 / 9 of the field of view, so the K value is used to represent the proportion of the target to the field of view, and it is expressed as the ratio of the target area to the market area.

[0042] The deviation calculation module calculates the deviation value of the result (x, y, K) detected by the target detection module;

[0043] According to the target center coordinates provided by the target detection module and the center position of the camera field of view, the module calculates the deviation values ​​in the horizontal and vertical directions in real time, and records the area size of the target in the image at the same time. That is, the deviation calculation module is used to calculate the offset between the center of the camera field of view and the center of the target, that is, the miss amount (detection output result (x, y, K)). The miss amount calculation result contains the precise position offset information of the target relative to the camera, which serves as the input parameter of the turntable adjustment module. Through precise deviation calculation, this module provides key control parameters for subsequent turntable adjustments, ensuring that the target can gradually be located in the center of the camera field of view and achieve maximum coverage.

[0044] The turntable adjustment module performs target alignment according to the off-target information transmitted by the deviation calculation module. In this embodiment, the control strategy of the turntable adjustment adopts a control algorithm based on PID (proportional-integral-differential). The PID control algorithm enables the camera to move quickly and smoothly to the center of the target through proportional, integral and differential control. In the specific implementation, proportional control is responsible for the adjustment of the current deviation, integral control is used to accumulate historical deviations, and differential control suppresses overshoot during the adjustment process. The turntable adjustment module continuously monitors and corrects the motion state of the turntable to avoid jitter caused by excessive adjustment, and gradually reduces the adjustment amplitude when the target is close to the center of the field of view, ensuring that the camera is finally accurately aligned with the target. The PID control algorithm ensures that the camera is quickly and accurately aligned with the target by adjusting the control parameters (such as the rotation speed and angle of the turntable) in real time to achieve the best target-seeking effect. The introduction of the PID control algorithm ensures the high response speed and adjustment accuracy of the system.

[0045] The feedback control module is used to continuously adjust and optimize the control parameters of the turntable according to the position information of the target in the current field of view during the process of the camera aligning the target, until the target is located in the center field of view of the camera and covers the entire field of view to the maximum extent. When the camera completes the accurate alignment of the target, the feedback control module will output a completion signal to mark the completion status of the current target in the multi-target system and enter the target search process for the next target.

[0046] In this embodiment, the feedback control process plays a role in the target-seeking task, ensuring that the target position is continuously in the center of the camera's field of view. The feedback control module plays a role in real-time adjustment in the entire target-seeking process. Specifically, the feedback control module receives status feedback from the turntable adjustment module and monitors the position change of the target position in the camera's field of view. If the target position does not reach the center of the field of view or does not maximize the coverage of the field of view, the feedback control module will dynamically update the control parameters to further optimize the adjustment of the turntable. When the camera is finally aimed at the target position and ensures that the target position is in the center of the field of view and maximizes coverage, the feedback control module outputs a completion signal, marking the completion of the current target position target-seeking task, and the system automatically enters the target-seeking process for the next target position. This module ensures that the system can continuously and stably achieve high-precision target seeking in dynamic and complex environments, especially in complex scenes with multiple targets and multiple cameras. Excellent performance.

[0047] like Figure 5 As shown in Figure 2, the workflow of the feedback control module is:

[0048] Step A: Set the N-1th target miss distance X n-1 , the N-1th turntable rotation amount Q n-1 ;

[0049] Step B: The Nth target miss distance is X n , the Nth turntable rotation angle is: Q n =X n* Q n-1 / (X n -X n-1 );

[0050] Step C: The N+1th target miss distance is X n+1 ;

[0051] Step D: The feedback control module determines whether the miss distance meets the requirements, that is, if |x|≤0.01, and |y|≤0.01, and Target search is completed; otherwise N=N+1, return to step A.

[0052] like Figure 7 As shown, the target-seeking system described in this embodiment is applied to target position recognition under different environmental conditions. The test results of the three images under different test backgrounds show that the system can stably identify and locate the target position, which directly demonstrates the excellent robustness and strong adaptability of the system.

[0053] Specific implementation method 2: Combination Figure 6 and Figure 7This embodiment is described as a method for intelligent target finding using a deep learning-based intelligent target finding system described in Specific Embodiment 1. This method uses a deep learning model to automatically identify and locate the target position, and combines the turntable control system to achieve precise alignment and maximize field of view coverage. The specific process of this method is as follows:

[0054] Step 1: Obtain high-resolution video frame images transmitted in real time by the camera through the data acquisition module;

[0055] Step 2: To improve the calculation speed of the neural network, the high-resolution video frame image obtained in step 1 is subjected to image preprocessing to obtain a low-resolution video frame image, and the RGB image is converted into a grayscale image;

[0056] Step 3: Using a feature extraction module to extract features from the low-resolution video frame image;

[0057] In this embodiment, a convolutional neural network (CNN) is used to perform multi-level feature extraction on the image, which can extract the edge, texture and shape features of the target image and obtain the target feature map;

[0058] Step 4: Use the target detection module to detect the coordinate position of the target in the image. After completing the target position recognition, the detection result is transmitted to the deviation calculation module. The deviation calculation module calculates the offset between the center of the camera field of view and the center of the target position according to the target position coordinates, that is, the miss distance;

[0059] Then, the turntable adjustment module is controlled in real time according to the miss distance information, and the two-dimensional turntable carried by the target camera is driven in real time by the turntable adjustment module to adjust the horizontal and pitch azimuths to ensure that the target position is always in the best field of view of the camera.

[0060] Step 5: The motion result verification module calculates whether the target center position and the field of view area occupied by the target meet the set requirements based on the deviation value after the turntable is adjusted, and performs feedback control based on the judgment result. If it does not meet the requirements, the target search step is re-executed from step 1; if it meets the requirements, the target search is completed;

[0061] Step 6: After target searching is completed, stop searching and the green light turns on to indicate that target searching is completed.

[0062] The technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above-mentioned embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0063] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. An intelligent target-seeking system based on deep learning, characterized by: The system includes a data acquisition module, a feature extraction module, a target detection module, a deviation calculation module, a turntable adjustment module and a feedback control module; The data acquisition module is used to obtain the real-time video frame image of each camera and transmit it to the feature extraction module; The feature extraction module extracts features from the video frame image, obtains edge, texture and shape features of the target image, and obtains a target object feature map; The target detection module uses a target detection algorithm based on deep learning to identify the target target feature map, obtains the coordinate position of the target target in the feature map, that is, the target center position; and transmits the target center position to the deviation calculation module; The deviation calculation module calculates the deviation value of the target center coordinate relative to the camera field of view center, i.e., the miss distance, according to the target center coordinate detected by the target detection module; The turntable adjustment module drives the turntable where the camera is located to adjust the direction and angle in real time according to the miss distance information obtained by the deviation calculation module, so as to ensure that the target position is always in the best field of view position of the camera; The feedback control module is used to continuously adjust and optimize the control parameters of the turntable according to the position information of the target in the current field of view during the process of the camera being aimed at the target, until the target is located in the center field of view of the camera and covers the entire field of view area to the maximum extent; When the camera completes accurate alignment with the target position, the feedback control module will output a completion signal.

2. The intelligent target-seeking system based on deep learning according to claim 1, characterized in that: The camera is mounted on a turntable and can freely adjust its direction and angle to adapt to different target positions.

3. The intelligent target-seeking system based on deep learning according to claim 1, characterized in that: The feature extraction module adopts the CNN network model to extract the edge, texture and shape features of the target image through multi-layer convolution layer, pooling layer and normalization layer operations.

4. The intelligent target-seeking system based on deep learning according to claim 1, characterized in that: The target detection module adopts a lightweight deep learning model, uses the target feature map extracted by the feature extraction module, and predicts the target center coordinates and bounding box coordinates through the bounding box regression algorithm and loss function convergence and transmits them to the deviation calculation module.

5. The intelligent target-seeking system based on deep learning according to claim 1, characterized in that: The deviation calculation module calculates the deviation values ​​in the horizontal and vertical directions in real time according to the target center coordinates sent by the target detection module and the center position of the camera field of view, and simultaneously records the area size of the target in the image.

6. The intelligent target-seeking system based on deep learning according to claim 1, characterized in that: The turntable adjustment module adopts a PID-based control algorithm to adjust the control parameters in real time so that the camera is aligned with the target center.

7. The intelligent target-seeking system based on deep learning according to claim 1, characterized in that: The feedback control module receives status feedback from the turntable adjustment module and monitors the position change of the target in the camera field of view; if the target does not reach the center of the field of view or does not maximize the coverage of the field of view, the feedback control module dynamically updates the control parameters to further optimize the adjustment of the turntable; when the camera is finally aimed at the target and ensures that the target is in the center of the field of view and maximizes coverage, the feedback control module outputs a completion signal, marking the completion of the current target search task, and the system automatically enters the target search process for the next target.

8. An intelligent target-finding method based on deep learning, characterized by: The specific steps of this method are: Step 1: Obtain high-resolution video frame images transmitted in real time by the camera through the data acquisition module; Step 2: Preprocess the high-resolution video frame image obtained in step 1 to obtain a low-resolution video frame image, and convert the RGB image into a grayscale image; Step 3: Use the CNN network model as a feature extraction module to perform feature extraction on the low-resolution video frame image, extract the edge, texture and shape features of the target image, and obtain a target feature map; Step 4: Use the target detection module to identify the target position; Detecting the coordinate position of the target object in the target object feature map, after completing the target identification, transmitting the target center coordinate to the deviation calculation module, and the deviation calculation module calculates the offset between the camera field of view center and the target center coordinate according to the target center coordinate, that is, the miss distance; Step 5, the deviation calculation module transmits the miss distance information to the turntable adjustment module, and the turntable adjustment module drives the turntable where the camera is located to adjust the direction and angle in real time according to the miss distance information to ensure that the target position is always in the best field of view position of the camera; When the camera is aimed at the target, the feedback control module continuously adjusts and optimizes the control parameters of the turntable according to the position information of the target in the current field of view, until the target is located in the center field of view of the camera and covers the entire field of view area to the maximum extent; When the camera completes accurate alignment with the target position, the feedback control module will output a completion signal.

9. The intelligent target-seeking method based on deep learning according to claim 8, characterized in that: In step 5, the workflow of the feedback control module is: Step A: Set the N-1th target miss distance X n-1 , the N-1th turntable rotation amount Q n-1 ; Step B: The Nth target miss distance is X n , the Nth turntable rotation angle is: Q n＝ X n* Q n-1 / (X n -X n-1 ); Step C: The N+1th target miss distance is X n+1 ; Step D: The feedback control module determines whether the miss distance (x, y, K) meets the requirements. If yes, the target search is completed; otherwise, n=n+1, and returns to step A.

10. The intelligent target-seeking method based on deep learning according to claim 9, characterized in that: If |x|≤0.01, and |y|≤0.01, and Target search completed.

Images