Infrared and visible light image fusion method and system based on target detection guidance

Abstract

The invention discloses an infrared and visible light image fusion method and system based on target detection guidance, and the method and system are applied to the technical field of image processing. Through the target extraction of an infrared image and a visible light image, on one hand, the multi-target fusion operation of the two images is realized, and the information complementation of the visible light and the infrared image is realized; on the one hand, the identification limitation of a target in a single-mode image is made up, on the other hand, whether the two images have the possibility of distortion is judged through the identification condition of a common target in the two images, and the type of the distorted image can be further judged by taking a variable-temperature target source preset in a scene as a standard target in an image distortion verification process. Therefore, the fusion weights of the two images are redistributed according to the distortion condition, and the image fusion accuracy is effectively improved.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and in particular to a method and system for infrared and visible light image fusion based on target detection guidance. Background Technology

[0002] Infrared and visible light images are two important types of optical images. Infrared images can reflect the thermal radiation information of objects and can effectively detect targets in harsh environments such as darkness and smoke, but they are insufficient in showing texture details. Visible light images have rich texture and detail information and conform to human visual habits, but their image quality is easily affected by environmental factors such as lighting and weather. Therefore, integrating the complementary information of these two types of images into a single image is an extremely important means of providing more comprehensive and reliable information and improving the accuracy of target detection.

[0003] For example, the infrared and visible light image fusion method with target feature embedding disclosed in Chinese patent CN116434024B solves the problem of mismatch between target detection tasks and image fusion tasks by constructing an infrared and visible light image fusion method with target feature embedding. This effectively improves the image fusion results for target detection tasks. Moreover, the target detection meta-feature embedding process is only used during training and does not add additional computation during the testing phase after training is completed. Another example is the target detection method with infrared and visible light image fusion disclosed in Chinese patent CN112560763B, which uses DenseBlock for feature extraction, solves the gradient vanishing problem of deep networks, and adopts an adaptive feature fusion strategy for different characteristics of extracted features.

[0004] However, most existing fusion methods focus on pixel-level or feature-level fusion, lacking targeted processing of specific semantic targets in images. They also ignore image errors caused by hardware failures, which seriously affect the fusion effect and the detection accuracy of targets. Summary of the Invention

[0005] The core of this invention lies in achieving an adaptive weighted fusion process for image features by extracting target features from infrared and visible light images and judging and verifying image distortion. This solves the problems of existing technologies, such as the lack of targeted processing for image targets and the inability to judge image distortion. At the same time, by giving the variable-temperature target source the mobility characteristic, infrared and visible light images of the variable-temperature target source at different positions are collected, which improves the accuracy of judging image distortion.

[0006] To solve the above problems, the present invention adopts the following technical solution.

[0007] A target detection-guided infrared and visible light image fusion method includes the following steps: S1. Select infrared and visible light images of the same scene Q collected at the same time, and extract targets from the two images separately, as follows: the same target detected in both infrared and visible light images is denoted as feature A, the target detected only in the infrared image is denoted as feature B, and the target detected only in the visible light image is denoted as feature C. Based on the bounding box information of features A and B in the infrared image, determine the background region M in the infrared image, and based on the bounding box information of features A and C in the visible light image, determine the background region N in the visible light image. S2. When feature A exists, assign equal weight coefficients to the infrared image and the visible light image, and proceed to step S3; when feature A does not exist but either feature B or feature C exists, proceed to step S4. S3. Perform weighted fusion on background region M and background region N to obtain fused background features. Perform weighted fusion on feature A, which is common to infrared and visible light images, to obtain fused common features. Merge common features, feature B, and feature C into background features to obtain fused image. S4. Image Verification: S4-1. Activate the variable temperature target source that was previously placed in scene Q, and assign the variable temperature target source a known temperature and color; S4-2. At the same time, acquire infrared and visible light images of scene Q; S4-3. Extract feature data q1 and feature data q2 about the variable temperature target source from the infrared image and visible light image obtained in step S4-2. Compare feature data q1 with the corresponding standard infrared image feature qb to obtain similarity Bb. Compare feature data q2 with the corresponding standard visible light image feature qc to obtain similarity Cc. S4-4. When Bb is higher than the set value B1 and Cc is lower than the set value C1, the visible light image is determined to be distorted. At this time, the infrared image is given a high weight coefficient and the visible light image is given a low weight coefficient. Then proceed to step S5. When Bb is lower than the set value B1 and Cc is higher than the set value C1, the infrared image is determined to be distorted. At this time, the infrared image is given a low weight coefficient and the visible light image is given a high weight coefficient, and then step S5 is performed. When Bb is higher than the set value B1 and Cc is higher than the set value C1, or when Bb is lower than the set value B1 and Cc is lower than the set value C1, the infrared image and the visible light image are assigned equal weight coefficients, and then step S5 is performed. Step S5: Perform weighted fusion on background region M and background region N to obtain fused background features. Merge features B and C into the background features to obtain the fused image.

[0008] Furthermore, the acquisition of standard infrared image features qb includes the following steps: under normal lighting and equipment conditions, and with the variable temperature target source enabled, a standard infrared image of scene Q is acquired, and feature data about the variable temperature target source in the standard infrared image is extracted as standard infrared image features qb. The acquisition of standard visible light image features qc includes the following steps: under normal lighting and equipment conditions, and with the variable temperature target source enabled, a standard visible light image of scene Q is acquired, and the feature data of the variable temperature target source in the standard visible light image is extracted as the standard visible light image features qc.

[0009] An infrared and visible light image fusion system based on target detection guidance is applied to the aforementioned infrared and visible light image fusion method based on target detection guidance. The system includes an infrared image acquisition module, a visible light image acquisition module, a feature extraction module, a weight allocation module, and a fusion module. The infrared image acquisition module acquires infrared images of scene Q through an infrared camera, and the visible light image acquisition module acquires visible light images of scene Q through a visible light camera.

[0010] Furthermore, the feature extraction module is used to extract target and background regions from infrared and visible light images, the weight allocation module is used to assign different weight coefficients to different fusion objects when fusing images, and the fusion module is used to fuse different objects.

[0011] Furthermore, the feature extraction module is connected to an image verification module, which is used to activate the variable temperature target source.

[0012] Furthermore, the image verification module is connected to a target feature database, which stores standard infrared image features qb and standard visible light image features qc.

[0013] Furthermore, the variable temperature target source includes a support base, a target column is fixedly connected to the upper end of the support base, and multiple evenly distributed strip-shaped holes are opened on the side end of the target column. A heat-conducting plate is fixedly connected inside the strip-shaped holes, and a heating wire is fixedly connected to the end face of the heat-conducting plate near the inner side of the target column.

[0014] Optionally, the lower end of the bearing seat is fixedly connected to a multi-section telescopic cylinder, the lower end of the multi-section telescopic cylinder is fixedly connected to a base, the upper end of the base is fixedly connected to a hydraulic cylinder, the movable end of the hydraulic cylinder is fixedly connected to the lower end of the bearing seat, and the hydraulic cylinder is located inside the multi-section telescopic cylinder.

[0015] Compared with the prior art, the advantages of this invention are: This solution extracts targets from infrared and visible light images. On the one hand, it enables multi-target fusion of the two types of images, achieving information complementarity between visible and infrared images and compensating for the limitations of target recognition in single-modal images. On the other hand, by identifying common targets in the two types of images, it determines whether there is potential for distortion in the two images. Furthermore, by using a pre-set variable-temperature target source in the scene as a standard target in the image distortion verification process, the type of distorted image can be further determined. Thus, the fusion weights of the two images can be redistributed according to the distortion situation, effectively improving the accuracy of image fusion. By making the target source movable, the position of the heat source formed by the heat-conducting plate and the heating wire can be changed. During the image verification process, not only can infrared and visible light images of the target source at different positions be acquired, but the accuracy of image acquisition is also improved, thereby improving the accuracy of judging the distortion of infrared and visible light images. Attached Figure Description

[0016] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a system block diagram of the present invention; Figure 3 Three-dimensional representation of the variable temperature target source of the present invention Figure 1 ; Figure 4 This is a schematic diagram of the front structure of the variable temperature target source of the present invention. Figure 1 ; Figure 5 Three-dimensional representation of the variable temperature target source of the present invention Figure 2 ; Figure 6 This is a schematic diagram of the front structure of the variable temperature target source of the present invention. Figure 2 ; Figure 7 This is a schematic diagram of the front structure of the variable temperature target source of the present invention. Figure 3 .

[0017] Explanation of the labels in the diagram: 1. Support base, 2. Target column, 3. Heat-conducting plate, 4. Heating wire, 5. Base, 6. Multi-section telescopic cylinder, 7. Hydraulic cylinder. Detailed Implementation

[0018] The technical solutions will now be clearly and completely described with reference to the accompanying drawings in the embodiments of the present invention.

[0019] First implementation method: Please see Figure 1An infrared and visible light image fusion system based on target detection guidance includes an infrared image acquisition module, a visible light image acquisition module, a feature extraction module, a weight allocation module, and a fusion module. The infrared image acquisition module acquires infrared images of scene Q through an infrared camera, and the visible light image acquisition module acquires visible light images of scene Q through a visible light camera.

[0020] The feature extraction module is used to extract target and background regions from infrared and visible light images, the weight allocation module is used to assign different weight coefficients to different fusion objects when fusing images, and the fusion module is used to fuse different objects.

[0021] A target detection-guided infrared and visible light image fusion method includes the following steps: S1. Select infrared and visible light images of the same scene Q collected at the same time, and extract targets from the two images separately, as follows: the same target detected in both infrared and visible light images is denoted as feature A, the target detected only in the infrared image is denoted as feature B, and the target detected only in the visible light image is denoted as feature C. Based on the bounding box information of features A and B in the infrared image, determine the background region M in the infrared image, and based on the bounding box information of features A and C in the visible light image, determine the background region N in the visible light image. S2. When feature A exists, assign equal weight coefficients to the infrared image and the visible light image, and proceed to step S3; when feature A does not exist but either feature B or feature C exists, proceed to step S4. S3. Perform weighted fusion on background region M and background region N to obtain fused background features. Perform weighted fusion on feature A, which is common to infrared and visible light images, to obtain fused common features. Merge common features, feature B, and feature C into background features to obtain fused image. When feature A is present, it indicates that the same target object was detected in both the infrared and visible light images, which to some extent shows that both image acquisition methods have achieved effective detection. In this case, the average weighting method is used to fuse the two. Conversely, when feature A is absent but feature B or feature C is present, this indicates that only one of the two images acquired at the same time for the same scene captured the target object, or the two images captured different target objects. This indicates that at least one of the two acquired images is distorted. To further determine the type of distorted image and improve the accuracy of image fusion, step S4 can be performed.

[0022] S4. Image Verification: S4-1. Activate the variable temperature target source that was previously placed in scene Q, and assign the variable temperature target source a known temperature and color; The temperature can be 50℃-55℃, which distinguishes it from the atmospheric temperature or room temperature under normal conditions, so that the variable temperature target source and the surrounding background area form different color gradients in the infrared image; the variable temperature target source can have its own color that is different from the background area, such as yellow, green, etc., and under normal lighting conditions, the variable temperature target source can be clearly imaged in the visible light image. S4-2. At the same time, acquire infrared and visible light images of scene Q; Since the variable temperature target source is in scene Q, it will be imaged in both infrared and visible light images. Note: Although the variable temperature target source will also be imaged in both infrared and visible light images in step S1, it is inactive at this time. Therefore, during feature extraction, the variable temperature target source can be ignored and not treated as the same target in both images. S4-3. Extract feature data q1 and feature data q2 about the variable temperature target source from the infrared image and visible light image obtained in step S4-2. Compare feature data q1 with the corresponding standard infrared image feature qb to obtain similarity Bb. Compare feature data q2 with the corresponding standard visible light image feature qc to obtain similarity Cc. The acquisition of standard infrared image feature qb includes the following steps: During the period from the completion of the infrared camera installation to its formal use, under the condition that the illumination and equipment are normal and the variable temperature target source is enabled (the temperature of the variable temperature target source is the same as the temperature in step S4-1), a standard infrared image of scene Q is acquired, and the feature data of the variable temperature target source in the standard infrared image is extracted as the standard infrared image feature qb. The acquisition of standard visible light image features qc includes the following steps: when the illumination and equipment are normal and the variable temperature target source is enabled (the temperature of the variable temperature target source is the same as the temperature in step S4-1), a standard visible light image of scene Q is acquired, and the feature data of the variable temperature target source in the standard visible light image is extracted as the standard visible light image features qc. Under normal circumstances (i.e., when neither the infrared camera nor the variable-temperature target source is faulty), the imaging state of the variable-temperature target source in the infrared image of step S4-2 is no different from its imaging state in the standard infrared image. This means that feature data q1 will normally be close to feature qb of the standard infrared image. If the similarity between the two is low, it indicates that the infrared image of step S4-2 may be distorted. Similarly, under normal circumstances (i.e., when neither the visible light camera nor the variable-temperature target source is faulty), the imaging state of the variable-temperature target source in the visible light image of step S4-2 is no different from its imaging state in the standard visible light image. This means that feature data q2 will normally be close to feature qc of the standard visible light image. If the similarity between the two is low, it indicates that the visible light image of step S4-2 may be distorted. Therefore, the similarity between feature data q1 and standard infrared image feature qb, and the similarity between feature data q2 and standard visible light image feature qc, can reflect the fault status of the infrared camera and the visible light camera respectively (without considering the special case of faults in the variable-temperature target source itself). Then, weighting coefficients are assigned based on the fault status of the two cameras, as follows: S4-4. When Bb is higher than the set value B1 and Cc is lower than the set value C1, the visible light image is determined to be distorted. At this time, the infrared image is given a high weight coefficient and the visible light image is given a low weight coefficient. Then proceed to step S5. When Bb is lower than the set value B1 and Cc is higher than the set value C1, the infrared image is determined to be distorted. At this time, the infrared image is given a low weight coefficient and the visible light image is given a high weight coefficient, and then step S5 is performed. When Bb is higher than the set value B1 and Cc is higher than the set value C1, or when Bb is lower than the set value B1 and Cc is lower than the set value C1, the infrared image and the visible light image are assigned equal weight coefficients, and then step S5 is performed. Step S5: Perform weighted fusion on background region M and background region N to obtain fused background features. Merge features B and C into the background features to obtain the fused image.

[0023] This application extracts targets from infrared and visible light images, enabling multi-target fusion of the two types of images. This achieves information complementarity between visible and infrared images, compensating for the limitations of target recognition in single-modal images. Furthermore, by identifying common targets in the two types of images, it determines whether there is potential for distortion. Moreover, by using a pre-set variable-temperature target source in the scene as a standard target in the image distortion verification process, the type of distorted image can be further determined. Thus, the fusion weights of the two images can be redistributed according to the distortion situation, effectively improving the accuracy of image fusion.

[0024] Please see Figure 2 and Figure 3 The variable temperature target source includes a support base 1, a target column 2 fixedly connected to the upper end of the support base 1, multiple evenly distributed strip holes on the side end of the target column 2, a heat-conducting plate 3 fixedly connected inside the strip holes, a heating wire 4 fixedly connected to the end face of the heat-conducting plate 3 near the inner side of the target column 2, an image verification module connected to the feature extraction module, the heating wire 4 connected to the image verification module, and a target feature database connected to the image verification module. The target feature database is used to store standard infrared image features qb and standard visible light image features qc.

[0025] When performing step S4-1, the heating wire 4 is energized through the image verification module, causing it to heat up. The heat from the heating wire 4 is directly transferred to the heat-conducting plate 3, gradually increasing the overall temperature of the heat-conducting plate 3. After the heating wire 4 has heated to a predetermined time (this time is preset by those skilled in the art), the heat-conducting plate 3 reaches a preset temperature of approximately 50°C. At this point, when the infrared camera acquires an image of the scene Q containing the variable-temperature target source, the color state of the variable-temperature target source in the infrared image will be clearly different from the surrounding background area. In addition, the outer surface of the variable-temperature target source has its own color and texture. Under sufficient lighting conditions, its color and texture information can be clearly displayed on the visible light image. This facilitates the acquisition of standard infrared image features qb and standard visible light image features qc before formal use, as well as the acquisition of actual imaging information during actual use, and the comparison of the two. This enables the analysis and judgment of hardware faults in the infrared and visible light cameras.

[0026] Second implementation method: This embodiment further refines the structure of the variable temperature target source based on the first embodiment, as detailed below: Please refer to... Figure 4 and Figure 5 The lower end of the support base 1 is fixedly connected to a multi-section telescopic cylinder 6, and the lower end of the multi-section telescopic cylinder 6 is fixedly connected to a base 5. The base 5 facilitates the placement of the variable temperature target source at a certain position in the scene Q. During installation, it is necessary to ensure that both the infrared camera and the visible light camera can fully image it and that it is not easily blocked by other objects in the scene. The upper end of the base 5 is fixedly connected to a hydraulic cylinder 7. The movable end of the hydraulic cylinder 7 is fixedly connected to the lower end of the support base 1, and the hydraulic cylinder 7 is located inside the multi-section telescopic cylinder 6. The hydraulic cylinder 7 is connected to the image verification module.

[0027] In the first embodiment, a static variable temperature target source is formed by the support 1, the target column 2, the heat-conducting plate 3 and the heating wire 4. After being heated, it can form different color distribution areas on the infrared image. However, since it is in a static state, the acquired image is easily affected by external factors (such as background noise, local contamination of the infrared camera lens, and blurring of the heat source outline caused by the heat-conducting plate 3 transferring heat to surrounding objects during the heating process), resulting in inaccuracies. Therefore, based on the above practical problems, this embodiment, through the arrangement of multi-section telescopic cylinder 6 and hydraulic cylinder 7, makes the heat source formed by heat-conducting plate 3 and heating wire 4 movable. This not only allows for the acquisition of infrared and visible light images of the variable-temperature target source at different positions, but also improves the accuracy of the images, thereby enhancing the accuracy of judging the distortion of infrared and visible light images. The specific operation is as follows: 1. Keep step S4-1 unchanged, that is, perform the same operation to heat up the heat-conducting plate 3; II. Regarding the single-point acquisition in step S4-2 of the first embodiment, this embodiment adopts the following acquisition operation instead: After the heating wire 4 is heated to the predetermined time, the infrared image acquisition module and the visible light image acquisition module are activated to acquire images of the temperature-changing target source (hydraulic cylinder 7 is in the retracted state, and target column 2 is at its lowest height) and scene Q in the initial state. Subsequently, the hydraulic cylinder 7 is activated intermittently to make the hydraulic cylinder 7 extend a fixed stroke once, so that the target column 2 stays at different height positions for a certain period of time. For example, if the initial height of the target column 2 from the ground is H1, and the fixed stroke of the hydraulic cylinder 7 in a single extension is h, then the height of the target column 2 from the ground during the intermittent stay is H1+h, H1+2h, H1+3h...H1+nH, where n is the number of times the hydraulic cylinder 7 extends, and the stay time can be 8-15 seconds. During the stay phase, the infrared image acquisition module and the visible light image acquisition module are activated to acquire images of the temperature-changing target source and scene Q after the height change, thereby obtaining multiple infrared images and visible light images. Feature data q1 of the variable-temperature target source in each infrared image and feature data q2 of the variable-temperature target source in each visible light image are extracted respectively. Each feature data q1 is compared with the corresponding standard infrared image feature qb to obtain a set of similarity Bb. Each feature data q2 is compared with the corresponding standard visible light image feature qc to obtain a set of similarity Cc. Note: The multiple standard infrared image features qb and multiple standard visible light image features qc are obtained using the same operation method as above during the period from camera installation to formal use. When the number of Bb values ​​exceeding the set value B1 accounts for half or more of the total number of Bb values, it indicates that the infrared image acquisition quality is normal. Conversely, if the number of Bb values ​​exceeding the set value B1 is less than half of the total number of Bb values, it indicates that the infrared image may be distorted. When the number of Cc values ​​exceeding the set value C1 accounts for half or more of the total number of Cc values, it indicates that the visible light image acquisition quality is normal. Conversely, if the number of Cc values ​​exceeding the set value C1 is less than half of the total number of Cc values, it indicates that the visible light image may be distorted.

[0028] The above description is merely a preferred embodiment of the present invention; it encompasses all the protection scope of the present invention. Any equivalent substitutions or modifications made by those skilled in the art within the technical scope disclosed in the present invention, based on the technical solutions and improved concepts of the present invention, should be covered within the protection scope of the present invention.

Claims

1. An infrared and visible image fusion method based on target detection guidance, characterized in that: The method comprises the following steps: S1, selecting infrared images and visible light images collected at the same time for the same scene Q, and extracting targets from the two kinds of images, specifically as follows: the same target detected in the infrared images and the visible light images is marked as feature A, the target detected only in the infrared images is marked as feature B, and the target detected only in the visible light images is marked as feature C; the background area M in the infrared images is determined according to the boundary box information of feature A and feature B in the infrared images; the background area N in the visible light images is determined according to the boundary box information of feature A and feature C in the visible light images; S2, when feature A exists, the infrared images and the visible light images are given equal weight coefficients, and step S3 is performed; when feature A does not exist but either feature B or feature C exists, step S4 is performed; S3, the background area M and the background area N are weighted and fused to obtain a fused background feature, the common feature A in the infrared images and the visible light images is weighted and fused to obtain a fused common feature, and the common feature, feature B and feature C are merged into the background feature to obtain a fused image; S4, image verification: S4-1, starting a variable-temperature target source pre-placed in the scene Q, and giving the variable-temperature target source a known temperature and color; S4-2, collecting infrared images and visible light images of the scene Q at the same time; S4-3, extracting feature data q1 and feature data q2 of the variable-temperature target source in the infrared images and the visible light images obtained in step S4-2, comparing the feature data q1 with the corresponding standard infrared image feature qb to obtain a similarity Bb, and comparing the feature data q2 with the corresponding standard visible light image feature qc to obtain a similarity Cc; S4-4, when Bb is higher than a set value B1 and Cc is lower than a set value C1, it is determined that the visible light images are distorted, at this time, the infrared images are given a high weight coefficient, the visible light images are given a low weight coefficient, and step S5 is performed again; when Bb is lower than a set value B1 and Cc is higher than a set value C1, it is determined that the infrared images are distorted, at this time, the infrared images are given a low weight coefficient, the visible light images are given a high weight coefficient, and step S5 is performed again; when Bb is higher than a set value B1 and Cc is higher than a set value C1, or Bb is lower than a set value B1 and Cc is lower than a set value C1, the infrared images and the visible light images are given equal weight coefficients, and step S5 is performed again; Step S5, the background area M and the background area N are weighted and fused to obtain a fused background feature, and feature B and feature C are merged into the background feature to obtain a fused image.

2. The method according to claim 1, wherein the method is characterized by: The acquisition operation of the standard infrared image feature qb comprises the following: under the condition that the light and the equipment are normal and the variable-temperature target source is enabled, a standard infrared image of the scene Q is collected, and the feature data of the variable-temperature target source in the standard infrared image is extracted as the standard infrared image feature qb; The obtaining operation of the standard visible light image feature qc includes the following: under the condition that the light and the device are normal, and the temperature-variable target source is enabled, a standard visible light image of the scene Q is collected, and feature data about the temperature-variable target source in the standard visible light image is extracted as the standard visible light image feature qc.

3. An infrared and visible image fusion system based on target detection guidance, applied to the infrared and visible image fusion method based on target detection guidance in claim 1, characterized in that: The infrared image acquisition module acquires an infrared image of the scene Q through an infrared camera, and the visible light image acquisition module acquires a visible light image of the scene Q through a visible light camera.

4. The infrared and visible image fusion system based on target detection guidance according to claim 3, characterized in that: The feature extraction module is used for extracting a target and a background region from the infrared image and the visible light image, the weight distribution module is used for distributing different weight coefficients to different fusion objects when the images are fused, and the fusion module is used for fusing different objects.

5. The infrared and visible image fusion system based on target detection guidance according to claim 3, characterized in that: The image verification module is connected to the feature extraction module, and is used for starting the temperature-variable target source.

6. The infrared and visible image fusion system based on target detection guidance according to claim 5, characterized in that: The image verification module is connected to the target feature database, and the target feature database stores the standard infrared image feature qb and the standard visible light image feature qc.

7. The infrared and visible image fusion system based on target detection guidance according to claim 5, characterized in that: The temperature-variable target source comprises a bearing seat (1), an upper end of the bearing seat (1) is fixedly connected with a target column (2), a side end of the target column (2) is provided with a plurality of evenly distributed strip-shaped holes, an inner portion of the strip-shaped hole is fixedly connected with a heat conduction plate (3), and an end face close to an inner side of the target column (2) of the heat conduction plate (3) is fixedly connected with a heating wire (4).

8. The infrared and visible image fusion system based on target detection guidance according to claim 7, characterized in that: A lower end of the bearing seat (1) is fixedly connected with a plurality of telescopic cylinders (6), a lower end of the plurality of telescopic cylinders (6) is fixedly connected with a base (5), an upper end of the base (5) is fixedly connected with a hydraulic cylinder (7), a movable end of the hydraulic cylinder (7) is fixedly connected with the lower end of the bearing seat (1), and the hydraulic cylinder (7) is located at an inner side of the plurality of telescopic cylinders (6).

Images