Image target recognition and guidance method and system based on five-axis turntable simulation

Abstract

The present invention discloses an image target recognition and guidance method and system based on five-axis turntable simulation, the method includes the inner three axes of the five-axis turntable simulating the flight attitude of the aircraft, the outer two axes of the five-axis turntable simulating the relative position relationship between the target and the aircraft; the infrared point source target simulator simulates the optical characteristics of the target at infinity, and the infrared target detector images the target optical characteristics simulated by the target simulator; the image processing computer obtains the image data of the target from the infrared target detector, identifies and tracks the target in the acquired image, locates its pixel coordinates in the image, and solves the guidance information of the target relative to the line of sight of the target detector in real time, and then sends the guidance information to the flight control computer, and the flight control computer guides the simulated aircraft in the simulation computer to fly to the target. The present invention can realize the optical guidance semi-physical hardware in-loop closed-loop simulation test based on five-axis turntable simulation, and has the advantage of effectively improving the simulation authenticity.

Description

Technical Field

[0001] The present invention relates to the technical field of target recognition, and in particular to an image target recognition and guidance method and system based on five-axis turntable simulation. Background Art

[0002] In the existing semi-physical simulation schemes for optically guided aircraft, a semi-physical simulation system scheme of a five-axis simulation turntable system or a three-axis turntable system plus a two-axis system is usually used. In the semi-physical simulation of optical guidance, physical objects such as target simulators and optically guided target detectors are usually introduced into the simulation loop, and some parts that are difficult to replace with physical objects are replaced by mathematical models and connected to the simulation loop. In semi-physical simulation, the credibility of the simulation is improved by replacing the real environment or equipment simulation with some physical objects, and many complex modeling problems in the past are also solved. This is the current main development direction of simulation experiments.

[0003] Image target recognition and tracking is one of the key technologies of optical guidance. Target recognition and tracking uses image processing algorithms to identify and locate the pixel coordinates of the target in the image, and then obtains the guidance information of the aircraft through coordinate conversion and solution. In traditional image target recognition, an artificially designed target feature extractor is used to extract and classify target features, among which support vector machine and k-nearest neighbor (KNN) algorithm are traditional pattern recognition methods. With the development of computer vision artificial intelligence, the feature information of the target is automatically extracted by deep learning neural network, and the CNN convolutional neural network is used to identify the target. In the target tracking scheme, the generative model of the single target tracking algorithm such as optical flow, meanshift algorithm, particle filter, Camshift tracking algorithm, etc. does not consider the background information. When the target is blocked, deformed, blurred or affected by light, it is easy to fail to track, and the generative tracking algorithm is inefficient. Compared with the generative tracking algorithm, discriminative tracking algorithms based on correlation filtering such as CSK algorithm, KCF algorithm, DSST algorithm, etc. can better obtain the specific position of the target in the current image, and have high recognition accuracy and fast tracking speed, so they are widely studied. With the development of deep learning, CNN convolutional neural network is used to extract image features, and the neural network is directly used to build an end-to-end twin network target tracking model.

[0004] In the existing semi-physical simulation system solution of a three-axis turntable system plus a two-axis system, the installation accuracy requirements of the three-axis turntable and the two-axis turntable are high, the mobility is poor, the flexibility of the semi-physical simulation test is limited, and the implementation of the solution is relatively difficult. Summary of the invention

[0005] In view of this, the present invention provides an image target recognition and guidance method and system based on five-axis turntable simulation.

[0006] The present invention discloses an image target recognition and guidance method based on five-axis turntable simulation, which comprises:

[0007] The inner three axes of the five-axis turntable simulate the flight attitude of the aircraft, and the outer two axes of the five-axis turntable simulate the positional relationship between the target and the aircraft in relative motion; the infrared point source target simulator is fixedly installed on the outer two axes of the five-axis turntable to simulate the optical characteristics of the target at infinity, and the infrared target detector is fixedly installed on the inner three axes of the five-axis simulation turntable to image the target optical characteristics simulated by the target simulator;

[0008] The image processing computer obtains the image data of the target from the infrared target detector, identifies, tracks and locates the pixel coordinates of the target in the acquired image, and calculates the guidance information of the target relative to the line of sight of the target detector in real time. The guidance information includes the misalignment angle and azimuth angle, and then sends the guidance information to the flight control computer. The flight control computer guides the simulated aircraft in the simulation computer to fly to the target.

[0009] Furthermore, the inner three axes of the five-axis turntable simulate the flight attitude of the aircraft during flight by rotating the pitch, yaw and roll angles, and the relative position relationship of the target relative to the line of sight of the aircraft is simulated by the inner three axes and the outer two axes of the five-axis turntable;

[0010] The infrared target simulator simulates the infrared point source target information. After the optical target detector is detected by the infrared target detector, the target image information obtained is a bright spot.

[0011] Further, the target image data is obtained from the optical target detector, and the target image data is grayed and binarized in turn to obtain a binary image; the connected area of ​​the binary image is obtained, and then the pixel area size of the connected domain is used for judgment to screen the point-like bright spot targets whose pixel area is within the preset threshold range as the screening result of the infrared point target; if multiple bright spot areas are screened out within the preset threshold range, the area with the largest area is selected from the multiple bright spot areas as the point target recognition area of ​​the current image, and the minimum enclosing rectangular frame of the point target recognition area of ​​the current image is calculated, and the minimum enclosing rectangular frame is recorded as (u, v, w, h) to complete the recognition and positioning of the infrared point source target in the image, where (u, v) is the center point coordinate of the pixel coordinate of the target positioned in the image, and (w, h) is the width and height of the target positioning frame; in the simulation process, after the target is recognized and positioned for the first time, it enters the tracking state, and continuously records the recognition and positioning coordinates of the target in the image of the historical preset frame before the current frame in subsequent frames, and updates the latest recorded target positioning coordinates in real time.

[0012] Furthermore, if the infrared point source target is identified and its positioning coordinates are obtained in the latest frame of the tracking phase, the average value of u of the target positioning point coordinates recorded in the historical continuous preset number of frames is calculated. and the average value of v , if the u value or v value of the target recognition position in the latest frame is equal to the average value of the historical continuous preset frames or If the absolute value of the difference is less than the preset value, the tracking and positioning coordinates (u, v) of the target are directly output; otherwise, the target position identified in the current frame is considered to be an interference wild point, and the target position identified in the previous frame is used as the tracking position of the target in the latest frame for output.

[0013] Furthermore, if the infrared point source target is not identified and its positioning coordinates are not determined in the latest frame of the tracking phase, it is determined whether the target is tracked and recorded. If the target is tracked and recorded and is in a tracked state, the target positioning coordinates retained in the previous frame are used as the output of the target position in the latest frame. If the target is in a tracking state and the target is not detected for a preset number of consecutive frames, the target is output as lost, the tracking state is exited, the retained historical target positioning coordinates are cleared, and the target positioning information is recognized again frame by frame.

[0014] Furthermore, the pixel coordinates (u, v) of the target in the image are converted into the sight misalignment angle and azimuth angle of the target in space to achieve flight guidance of the aircraft;

[0015] In the simulation process, the target guidance information is solved by outputting the pixel coordinates p(u,v) of the target recognition, tracking and positioning information in the image frame by frame, and then calculating the physical coordinates p(x,y) of the target in the imaging plane of the optical target detector through the pixel coordinates p(u,v). The calculation formula is:

[0016]

[0017] Among them, the unit of the principal point coordinate (u0, v0) is pixel, which represents the pixel position of the origin of the physical coordinate system of the image plane in the pixel coordinate system; dx and dy are scale factors, which are internal parameters of the optical target detector camera and are used to describe its imaging characteristics and geometric properties.

[0018] Furthermore, the positional relationship between the target pixel p(u, v) coordinates of the optical target detection imaging and the target positioning p(x, y) coordinates in the imaging plane of the optical target detector is a translation and rotation transformation relationship of the plane coordinate system.

[0019] Furthermore, in the process of solving the problem, the pixel coordinates (u, v) of the target obtained by target recognition and tracking are used to calculate the physical coordinates p (x, y) of the target in the imaging plane of the optical target detector, and then the focal length f of the optical target detector is used to calculate the misalignment angle in the optical target detection coordinate system. and azimuth angle θ; and the misalignment angle and azimuth angle θ are sent to the flight control computer to generate control command information, thereby guiding the simulated aircraft to fly to the target.

[0020] Furthermore, the calculation of the misalignment angle in the optical target detection coordinate system is and azimuth angle θ, specifically:

[0021]

[0022]

[0023]

[0024] Misalignment angle of target sight is the angle between the line connecting the aircraft and the target and the Z axis, and the azimuth angle is the angle θ between the projection of the target onto the XOY plane of the optical target detector coordinate system and the positive direction of the Z axis in a counterclockwise direction;

[0025] The misalignment angle and azimuth angle can be used to uniquely determine the angle guidance information of the target direction, which is converted from the pixel position p(u,v) of the target to the target point p in space. c The transformation relationship of (x, y, z) is the mapping transformation relationship from the two-dimensional coordinate system to the target direction point of the three-dimensional space coordinate system.

[0026] The present invention also discloses an image target recognition and guidance system based on five-axis turntable simulation, which is used to implement any of the above-mentioned image target recognition and guidance methods based on five-axis turntable simulation, and comprises:

[0027] A five-axis turntable, an infrared point source target simulator, an infrared target detector, an image processing computer, a flight control computer and a simulation computer; the inner three axes of the five-axis turntable are used to simulate the flight attitude of the aircraft, and the outer two axes of the five-axis turntable are used to simulate the positional relationship between the target and the aircraft in relative motion; the infrared point source target simulator is fixedly mounted on the outer two axes of the five-axis turntable to simulate the optical characteristics of the target at infinity, and the infrared target detector is fixedly mounted on the inner three axes of the five-axis simulation turntable to image the target optical characteristics simulated by the target simulator;

[0028] The optical target detector is connected to the flight control computer through the image processing computer; the flight control computer is connected to the five-axis turntable through the simulation computer; the image processing computer is used to obtain the image data of the target from the target detector, identify and track the target in the acquired image, locate its pixel coordinates, and calculate the guidance information of the target relative to the line of sight of the target detector in real time. The guidance information includes the misalignment angle and azimuth, and then send the guidance information to the flight control computer, which is used to guide the simulated aircraft in the simulation computer to fly to the target.

[0029] Due to the adoption of the above technical scheme, the present invention has the following advantages: a five-axis turntable is used to simulate the relative position relationship between the aircraft and the target, which is easy to operate and has high control accuracy. The target detector and part of the target simulator are introduced into the semi-physical simulation loop through the five-axis turntable, which can improve the credibility of the simulation; through the image target recognition algorithm, the position of the infrared target point source target in the image is automatically identified and locked, and the optical target detector identifies and tracks the target to obtain the line of sight misalignment angle and azimuth angle information, which are sent to the flight control computer through the serial port to automatically guide the aircraft without manual selection of the locked target. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the embodiments of the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0031] Figure 1 This is a structural diagram of an image target recognition and guidance system based on five-axis turntable simulation according to an embodiment of the present invention;

[0032] Figure 2 A schematic diagram of a flow chart of infrared target point source recognition, tracking and guidance information solution in an image processing computer according to an embodiment of the present invention;

[0033] Figure 3 A schematic diagram of the positional relationship between the target pixel coordinates and the positioning coordinates of the target on the image plane in the optical target detection imaging according to an embodiment of the present invention;

[0034] Figure 4 The figure is a schematic diagram of the positional relationship between the target pixel coordinates of the optical target detection imaging and the actual line of sight position coordinates of the target in the optical target detector according to an embodiment of the present invention. DETAILED DESCRIPTION

[0035] The present invention is further described in conjunction with the accompanying drawings and embodiments, and the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by ordinary technicians in this field should fall within the scope of protection of the embodiments of the present invention.

[0036] The present invention provides an embodiment of an image target recognition and guidance method based on five-axis turntable simulation, which comprises:

[0037] The inner three axes of the five-axis turntable simulate the flight attitude of the aircraft, and the outer two axes of the five-axis turntable simulate the positional relationship between the target and the aircraft in relative motion; the infrared point source target simulator is fixedly installed on the outer two axes of the five-axis turntable to simulate the optical characteristics of the target at infinity, and the infrared target detector is fixedly installed on the inner three axes of the five-axis simulation turntable to image the target optical characteristics simulated by the target simulator;

[0038] The image processing computer provides computing services, obtains the target's image data from the infrared target detector, identifies and tracks the target in the acquired image, locates its pixel coordinates, and calculates in real time the guidance information of the target relative to the target detector's line of sight. The guidance information includes the misalignment angle and azimuth angle, and then sends the guidance information to the flight control computer. The flight control computer guides the simulated aircraft in the simulation computer to fly to the target.

[0039] The inner three axes of the five-axis turntable simulate the flight attitude of the aircraft during flight by rotating the pitch, yaw and roll angles, and the relative position relationship of the target relative to the aircraft's line of sight is simulated by the inner three axes and the outer two axes of the five-axis turntable.

[0040] The image processing computer is connected to the infrared target detector of the optical guidance head, collects image information of the infrared target point source from the infrared target detector, and provides computing services to identify, track and locate the pixel coordinates of the target in the image, and calculates the misalignment angle and azimuth information of the target relative to the line of sight of the target detector in real time, and at the same time sends the identified and calculated target guidance information to the flight control computer to realize the optical guidance simulation of the simulated aircraft flying to the target.

[0041] In the simulation, the infrared target simulator simulates the infrared point source target information. After the optical target detector detects the infrared target detector, the target image information obtained is a bright spot, which is a typical spot target. In the two-dimensional image, there is an obvious grayscale difference between the target and the surrounding. The solution process of infrared target point source recognition, tracking and guidance information in the image processing computer is as follows Figure 2 shown.

[0042] First, the target image data of size 640×512 is obtained from the optical target detector, and then the obtained image is grayed so that each pixel value ranges from 0 to 255; then the obtained grayed image is binarized by setting a certain threshold method, and the pixel value above the threshold is 1 and the pixel value below the threshold is 0, so as to obtain a binary image of size 640×512. Then, the connected area of ​​the binary image is obtained by analyzing the neighboring connected domain of the image pixels. Finally, the pixel area size threshold of the connected domain is used to judge and filter the point-like bright spot targets with a pixel area within a certain threshold range as the screening result of the infrared point target. If multiple bright spot areas are screened out within the threshold range, the area with the largest area within the set threshold range is selected as the point target recognition area of ​​the current image. Finally, the minimum enclosing rectangular frame (u, v, w, h) of the screened bright spot area is calculated to complete the recognition and positioning of the infrared point source target in the image, where (u, v) is the center point of the target in the image positioning, and (w, h) is the width and height of the target positioning frame. After the target is identified and located for the first time during the simulation process, it enters the tracking state. In the subsequent frames, the identification and positioning coordinates of the target in the image of the previous 6 frames are continuously recorded, and the latest recorded target positioning coordinates are updated in real time.

[0043] If the infrared point source target is identified and located in the latest frame of the tracking phase, the average value of u of the target location point coordinates recorded in the past 6 consecutive frames is calculated. and the average value of v , if the u value or v value of the target recognition position in the latest frame is the same as the average value of the six consecutive frames in history and If the absolute value of the difference is less than a certain threshold, the target tracking and positioning coordinates (u, v) are directly output. If it is greater than, the target position identified in the current frame is considered to be a flying point, and the target position identified in the previous frame is used as the tracking position of the target in the latest frame for output. If the infrared point source target in the latest frame is not identified and positioned, it is determined whether the target is tracked and recorded. If the target is tracked and recorded in the tracked state, the target positioning coordinates retained in the previous frame are used as the output of the target position in the latest frame. If the target is in the tracking state and the target is not detected for a certain number of consecutive frames, the target is lost, the tracking state is exited, the retained historical target positioning coordinates are cleared, and the target positioning information is re-identified frame by frame.

[0044] In the simulation, the pixel coordinates (u, v) of the target in the image are located through image target recognition. The image (u, v) pixel coordinate information cannot be directly used for the flight guidance of the aircraft. The positioning coordinates of the target in the image need to be converted into the line of sight misalignment angle and azimuth information of the target in space in order to complete the flight guidance of the aircraft through image target recognition.

[0045] In the simulation process, the target guidance information is solved by outputting the pixel coordinates p(u,v) of the target recognition, tracking and positioning information in the image frame by frame. Then the physical coordinates p(x,y) of the target in the image plane of the optical target detector are calculated by the pixel coordinates p(u,v). The calculation formula is as follows.

[0046]

[0047] The unit of the principal point coordinates (u0, v0) is pixel (pix), which indicates the pixel position of the origin of the physical coordinate system of the image plane in the pixel coordinate system; dx and dy are scale factors with the unit of pixel value per micron (pix / um), which are internal parameters of the optical target detector camera and are used to describe its imaging characteristics and geometric properties.

[0048] The positional relationship between the target pixel p(u,v) coordinates of the optical target detection imaging and the target positioning p(x,y) coordinates on the image plane is the translation and rotation transformation relationship of the plane coordinate system, such as Figure 3 shown. Figure 3 In the figure, the origin of the similar coordinate system is in the upper left corner, the U axis is the horizontal axis of the similar coordinate system, and the V axis is the vertical axis of the pixel coordinate system.

[0049] In the process of solving, the pixel coordinates (u, v) of the target obtained by target recognition and tracking are used to calculate the physical coordinates p (x, y) of the target point on the image plane, and then the focal length f of the optical target detector is used to calculate the misalignment angle in the optical target detection coordinate system. and azimuth angle θ, which are calculated as follows.

[0050]

[0051]

[0052]

[0053] Misalignment angle of target sight The angle between the line of sight between the aircraft and the target and the Z axis. The azimuth angle is the angle θ between the target projected onto the XOY plane of the optical target detector coordinate system and the positive direction of the X axis in the counterclockwise direction. The misalignment angle and azimuth angle can be used to uniquely determine the target direction. The target pixel p(u,v) coordinates of the optical target detection image and the actual line of sight position of the target in the optical target detector are p c (x,y,z) The positional relationship between the coordinates is the mapping transformation relationship from the two-dimensional coordinate system to the target direction point of the three-dimensional space coordinate system, such as Figure 4 shown. Figure 4The optical target detector coordinate system (O-XYZ) is a right-handed coordinate system in space, with the detector optical center O as the coordinate origin and the optical axis outward as the Z axis. Any spatial coordinate point is p c (x,y,z) .

[0054] Image recognition is used to track the target and locate the coordinates (u, v) in the image pixels. The pixel coordinates of the target are solved into the misalignment angle and azimuth angle under the target line of sight through the solution of guidance information. The guidance information of the two angles is sent to the flight control computer through serial port information to generate control command information, thereby guiding the simulated aircraft to fly to the target.

[0055] See also Figure 1 The present invention also provides an embodiment of an image target recognition and guidance system based on five-axis turntable simulation, which is used to implement the image target recognition and guidance method based on five-axis turntable simulation described in the above embodiment, and includes:

[0056] A five-axis turntable, an infrared point source target simulator, an infrared target detector, an image processing computer, a flight control computer and a simulation computer; the inner three axes of the five-axis turntable are used to simulate the flight attitude of the aircraft, and the outer two axes of the five-axis turntable are used to simulate the positional relationship between the target and the aircraft in relative motion; the infrared point source target simulator is fixedly mounted on the outer two axes of the five-axis turntable to simulate the optical characteristics of the target at infinity, and the infrared target detector is fixedly mounted on the inner three axes of the five-axis simulation turntable to image the target optical characteristics simulated by the target simulator;

[0057] The optical target detector is connected to the flight control computer through the image processing computer; the flight control computer is connected to the five-axis turntable through the simulation computer; the image processing computer is used to provide computing services, obtain the image data of the target from the infrared target detector, identify and track the target in the acquired image, locate its pixel coordinates, and calculate the guidance information of the target relative to the line of sight of the target detector in real time. The guidance information includes the misalignment angle and azimuth, and then send the guidance information to the flight control computer. The flight control computer is used to guide the simulated aircraft in the simulation computer to fly to the target.

[0058] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. An image target recognition and guidance method based on five-axis turntable simulation, characterized in that: include: The inner three axes of the five-axis turntable simulate the flight attitude of the aircraft, and the outer two axes of the five-axis turntable simulate the positional relationship between the target and the aircraft in relative motion; The infrared point source target simulator is fixedly installed on the outer two axes of the five-axis turntable to simulate the optical characteristics of the target at infinity. The infrared target detector is fixedly installed on the inner three axes of the five-axis simulation turntable to image the target optical characteristics simulated by the target simulator. The image processing computer obtains the image data of the target from the infrared target detector, identifies and tracks the target in the obtained image, locates its pixel coordinates in the image, and calculates the guidance information of the target relative to the sight line of the target detector in real time. The guidance information includes two angle information of the misalignment angle and azimuth angle of the target sight line direction, and then sends the guidance information to the flight control computer through the serial port. The flight control computer calculates the action instructions according to the guidance information to guide the simulated aircraft in the simulation computer to fly to the target; Obtain target image data from the optical target detector, grayscale and binarize the target image data in sequence to obtain a binary image; obtain the connected area of ​​the binary image, and then use the pixel area size of the connected area to make a judgment, and screen out point-shaped bright spot targets whose pixel area is within a preset threshold range as the preliminary screening result of infrared point targets; If multiple bright spot areas are screened out within the preset threshold range, the area with the largest area is selected from the multiple bright spot areas as the point target recognition area of ​​the current image, and then the minimum enclosing rectangular frame of the point target recognition area of ​​the current image is calculated. The minimum enclosing rectangular frame is recorded as (u, v, w, h) to complete the recognition and positioning of the infrared point source target in the image; where (u, v) is the center point coordinate of the pixel coordinates of the target in the image, and (w, h) is the width and height of the target positioning frame; after the target is recognized and positioned for the first time during the simulation process, it enters the tracking state, and continuously records the recognition and positioning coordinate results of the target in the image for the preset number of frames in the history before the current frame in subsequent frames, and updates the latest recorded target positioning coordinates in real time; The pixel coordinates (u, v) of the target in the image are converted into two angle information: the misalignment angle and the azimuth angle of the target's line of sight in space, so as to achieve flight guidance for the aircraft; In the simulation process, the target guidance information is solved by outputting the pixel coordinates p(u,v) of the target recognition, tracking and positioning information in the image frame by frame, and then calculating the physical coordinates p(x,y) of the target in the imaging plane of the optical target detector through the pixel coordinates p(u,v). The calculation formula is: Among them, the unit of the principal point coordinates (u0, v0) is pixel, which represents the pixel position of the origin of the physical coordinate system of the image plane in the pixel coordinate system; dx and dy are scale factors, which are internal parameters of the optical target detector camera and are used to describe its imaging characteristics and geometric properties; In the process of solving, the pixel coordinates (u, v) of the target obtained by target recognition and tracking are used to calculate the physical coordinates p (x, y) of the target in the imaging plane of the optical target detector, and then the focal length f of the optical target detector is used to calculate the misalignment angle in the optical target detection coordinate system. and azimuth angle θ; and the misalignment angle and azimuth angle θ are sent to the flight control computer to generate control instruction information, thereby guiding the simulated aircraft to fly to the target; The calculation of the misalignment angle in the optical target detection coordinate system and azimuth angle θ, specifically: The misalignment angle is the angle between the line of sight between the aircraft and the target and the Z axis. The azimuth angle is the angle θ between the target projected onto the XOY plane of the optical target detector coordinate system, with the positive direction of the X axis counterclockwise as positive. The misalignment angle and azimuth angle can uniquely determine the angle guidance information of the target direction, and the pixel position p(u,v) of the target is converted to the target point p in space. c The transformation relationship of (x, y, z) is the mapping transformation relationship from the two-dimensional coordinate system to the target direction point of the three-dimensional space coordinate system.

2. The image target recognition and guidance method based on five-axis turntable simulation according to claim 1 is characterized in that: The inner three axes of the five-axis turntable simulate the flight attitude of the aircraft during flight by rotating the pitch, yaw and roll angles, and the relative position relationship of the target relative to the aircraft's line of sight is simulated by the inner three axes and the outer two axes of the five-axis turntable; The infrared target simulator simulates the infrared point source target information. After the optical target detector is detected by the infrared target detector, the target image information obtained is a bright spot.

3. The image target recognition and guidance method based on five-axis turntable simulation according to claim 1 is characterized in that: If the infrared point source target is identified and located in the latest frame of the tracking phase, the average value of the target location point coordinates u recorded in the historical continuous preset frame number is calculated. and the average value of v , if the u value or v value of the target recognition position in the latest frame is the average value of the preset number of frames recorded in the history or If the absolute value of the difference is less than a certain preset value, the tracking and positioning coordinates (u, v) of the target in the current frame are directly output; otherwise, the target position identified in the current frame is considered to be an interference wild point, and the target recognition and positioning result of the previous frame recorded in the past is used as the tracking and positioning result of the target in the current latest frame.

4. The image target recognition and guidance method based on five-axis turntable simulation according to claim 1 is characterized in that: If the infrared point source target is not identified and its positioning coordinates are not found in the latest frame of the tracking phase, it is determined whether the target is tracked and recorded. If the target is tracked and recorded in the tracked state, the target positioning coordinates recorded in the previous frame are used as the output of the target position in the latest frame. If the target is in the tracking state and the target is not detected for a preset number of consecutive frames, the target is lost, the tracking state is exited, the retained historical target positioning coordinates are cleared, and the target positioning information is recognized again frame by frame.

5. The image target recognition and guidance method based on five-axis turntable simulation according to claim 1 is characterized in that: The positional relationship between the target pixel p(u,v) coordinates of the optical target detection imaging and the target positioning p(x,y) coordinates in the imaging plane of the optical target detector is the translation and rotation transformation relationship of the plane coordinate system.

6. An image target recognition and guidance system based on five-axis turntable simulation, used to implement the image target recognition and guidance method based on five-axis turntable simulation according to any one of claims 1 to 5, characterized in that: include: A five-axis turntable, an infrared point source target simulator, an infrared target detector, an image processing computer, a flight control computer and a simulation computer; the inner three axes of the five-axis turntable are used to simulate the flight attitude of the aircraft, and the outer two axes of the five-axis turntable are used to simulate the positional relationship between the target and the aircraft in relative motion; the infrared point source target simulator is fixedly mounted on the outer two axes of the five-axis turntable to simulate the optical characteristics of the target at infinity, and the infrared target detector is fixedly mounted on the inner three axes of the five-axis simulation turntable to image the target optical characteristics simulated by the target simulator; The optical target detector is connected to the flight control computer through the image processing computer; the flight control computer is connected to the five-axis turntable through the simulation computer; the image processing computer is used to obtain the image data of the target from the target detector, identify and track the target in the acquired image, locate its pixel coordinates, and calculate the guidance information of the target relative to the line of sight of the target detector in real time. The guidance information includes the misalignment angle and azimuth, and then send the guidance information to the flight control computer, which is used to guide the simulated aircraft in the simulation computer to fly to the target.

Images