Alarm picture generation method and system based on multi-focal-length multi-view camera group

Abstract

The invention discloses an alarm picture generation method based on a multi-focal-length multi-view camera group, the multi-focal-length multi-view camera group comprises a first distance measuring camera and a second distance measuring camera, the focal length of the first distance measuring camera is smaller than that of the second distance measuring camera, and the method comprises the following steps: S1, shooting a preset area, and analyzing shot image information; s2, the first distance measuring camera or the second distance measuring camera detects a target, detected target alarm information is superposed on image information of the first distance measuring camera, and an alarm picture is obtained; and S3-S5, when the first distance measuring camera and the second distance measuring camera detect the same target, according to the pixel number of the target in the image information of the first distance measuring camera and by combining whether the target is complete in the image information of the second distance measuring camera, selecting the target alarm information of the first distance measuring camera or the second distance measuring camera to be superposed on the image information of the first distance measuring camera, and obtaining an alarm picture. The invention further discloses an alarm picture generation system which is used for executing the steps in the alarm picture generation method.

Description

Technical Field

[0001] The present invention relates to the field of intelligent monitoring technology, and in particular to a method and system for generating warning images based on a multi-focal-segment multi-eye camera group. Background Art

[0002] As we all know, with the widespread application of intelligent monitoring systems in security, traffic management and other fields, target detection and ranging technology based on camera visual recognition has become the core means to achieve intelligent perception, especially in high-voltage power grid scenarios. Cameras are often used to intelligently monitor target scenes, and after determining that a target appears in the target scene, the camera is used to capture and measure the distance. After the measurement is completed, the ranging alarm picture is directly uploaded to facilitate users to obtain the ranging results and avoid safety problems.

[0003] The inventors have found that in the current real-time monitoring system for power transmission lines, after using a multi-focal-segment multi-eye camera to capture and measure distance, the corresponding generated warning pictures usually have the following problems: In actual application, if the ranging warning picture detected by the telephoto rangefinder camera is uploaded directly, it is very likely that the captured picture cannot cover the entire scene due to the small field of view of the telephoto lens. The user cannot obtain the global position and size information of the target in the scene from the picture. Especially when the target is close to key objects such as wires, the wires may not be clearly displayed in the picture, and the shortest distance from the target to the wires cannot be accurately identified, which seriously affects the user's judgment of the target position and potential risks.

[0004] In addition, when the same target appears in the field of view of cameras with different focal lengths, multiple warning images will be generated and uploaded. Since cameras with different focal lengths shoot at different angles and distances, the shortest distance from the target to the wire in multiple images may be inconsistent. This not only increases the data processing burden of the system, but may also cause confusion in the user's judgment of the target distance, affecting the accuracy of the decision.

[0005] If the same target appears in cameras with different focal lengths, two or more warning images will be uploaded. It is possible that the shortest distance between the same target and the wire in the two warning images is inconsistent.

[0006] It can be seen that the disadvantages of the existing technical solutions are that the field of view is limited and it is impossible for users to obtain the distance measurement results easily and intuitively, which easily leads to confusion. Summary of the invention

[0007] The technical problem to be solved by the present invention is: the present invention discloses a method and system for generating a warning image based on a multi-focal-segment multi-eye camera group, so as to solve the problems of field of view limitation and target repeated warning existing in the warning image generated after the multi-focal-segment multi-eye camera is used to capture and measure the distance in the prior art solution.

[0008] In order to solve the above technical problems, the technical solution adopted by the present invention is: a method for generating a warning picture based on a multi-focal-segment multi-eye camera group, wherein the multi-focal-segment multi-eye camera group includes a first ranging camera and a second ranging camera, and the focal length of the first ranging camera is smaller than that of the second ranging camera, and the method includes the steps of: S1: photographing a preset area using a multi-focal-length multi-eye camera group, and analyzing the captured image information, so as to execute S2 or S3 when at least one ranging camera detects a target; S2: Only the first ranging camera or the second ranging camera detects the target, and directly superimposes the detected target warning information on the image information of the first ranging camera to obtain a warning picture; S3: If the first ranging camera and the second ranging camera detect the same target, the number of pixels of the target in the image information of the first ranging camera is obtained, and S4 or S5 is executed; S4: the number of pixels is ≥ the preset pixel number threshold, directly superimposing the target warning information detected by the first ranging camera on the image information of the first ranging camera to obtain a warning picture; S5: If the number of pixels is less than the preset pixel number threshold, it is determined whether the target is complete in the image information of the second ranging camera; if it is complete, the target warning information detected by the second ranging camera is superimposed on the image information of the first ranging camera to obtain a warning picture; if it is incomplete, the target warning information detected by the first ranging camera is superimposed on the image information of the first ranging camera to obtain a warning picture.

[0009] In the above technical solution of the present invention, the present invention discloses a new warning image generation method based on a multi-focal-segment multi-eye camera group, which can shoot a preset area through a first detection camera and a second ranging camera, and analyze the captured image information to generate warning images using different strategies according to different detection situations.

[0010] Specifically, in actual application, the focal length of the first ranging camera is smaller than that of the second ranging camera, and the field of view of the first ranging camera is larger. No matter what the detection situation is, the target warning information finally detected is superimposed on the image information of the first ranging camera to obtain a warning picture, thereby ensuring that the field of view of the warning picture is maximized in any case, so that the user can see the ranging result more intuitively.

[0011] At the same time, when the first and second ranging cameras detect the same target, in order to optimize the display effect of the warning picture obtained by the user, the warning picture generation method designed by the present invention can judge and select the target warning information of the appropriate ranging camera according to the number of pixels of the target in the image information of the first ranging camera and whether the target is complete in the image information of the second ranging camera, so as to superimpose it on the image information of the first ranging camera and obtain the warning picture, which has good promotion prospects and application value.

[0012] Furthermore, in the method for generating a warning image based on a multi-focal-length multi-eye camera group described in the present invention, the multi-focal-length multi-eye camera group includes a main device and a sub-device, and the main device and the sub-device both include a first ranging camera and a second ranging camera; wherein, in step S1, the first ranging camera and the second ranging camera in the main device and the sub-device are used simultaneously to shoot a preset area.

[0013] Furthermore, in the method for generating a warning picture based on a multi-focal-segment multi-eye camera group described in the present invention, in steps S2, S4 and S5, the detected target warning information is superimposed on the image information of the first ranging camera of the master device to obtain a warning picture.

[0014] In the above technical solution of the present invention, according to the actual application scenario, considering that a single device combines the first ranging camera and the second ranging camera for snapshot ranging, there are still certain usage defects; for this reason, in actual use, a multi-focal-segment multi-eye camera group can also be provided including a main device and a sub-device, and the main device and the sub-device are respectively provided at two different positions of the transmission tower, and the main device and the sub-device are controlled to include the first ranging camera and the second ranging camera, so as to perform detection ranging respectively.

[0015] Of course, in order to avoid the problem of user confusion caused by uploading multiple alarm pictures in the prior art, in actual application, no matter what the situation is, which camera detects the target alarm information, it is finally superimposed on the image information of the first ranging camera of the main device to obtain an alarm picture, so that the power grid staff can obtain the ranging results intuitively and accurately.

[0016] Furthermore, in the method for generating warning images based on a multi-focal-segment multi-camera group according to the present invention, the target warning information includes: a target segmentation result, a ranging line and a ranging value.

[0017] Furthermore, in the method for generating a warning image based on a multi-focal-segment multi-camera group according to the present invention, in step S2 or step S5, the target warning information detected by the second ranging camera is superimposed on the image information of the first ranging camera, specifically: According to the parameters of the second ranging camera and the detected target segmentation result, the target is projected into the three-dimensional space; and according to the parameters of the first ranging camera, the three-dimensional space target is projected onto the image information of the first ranging camera; According to the parameters of the second ranging camera, the two end points of the ranging line detected by the second ranging camera are projected into the three-dimensional space; and according to the parameters of the first ranging camera, the three-dimensional space end points are projected onto the image information of the first ranging camera; The distance measurement value detected by the second distance measurement camera is directly superimposed on the image information of the first distance measurement camera.

[0018] Further, in the method for generating a warning image based on a multi-focal-segment multi-camera group according to the present invention, in step S3, it is determined whether the first ranging camera and the second ranging camera detect the same target, specifically: According to the parameters of the first ranging camera and the detected target segmentation result, the target is projected into a three-dimensional space to obtain a first target three-dimensional point cloud set; According to the parameters of the first ranging camera and the detected target segmentation result, the target is projected into a three-dimensional space to obtain a second target three-dimensional point cloud set; The overlap between the first target three-dimensional point cloud set and the second target three-dimensional point cloud set is compared, and when the overlap reaches a preset overlap, it is determined that the first ranging camera and the second ranging camera have detected the same target.

[0019] In the above technical solution of the present invention, when the first detection camera and the second detection camera are used to monitor the preset area at the same time, it is inevitable that the first detection camera and the second detection camera will detect the target at the same time. At this time, it can be further determined whether the targets detected by the two are the same target. The targets detected by the first detection camera and the second detection camera can be projected into three-dimensional space according to the coordinate mapping relationship of the pixel points on the two cameras, and matched, so as to make a judgment based on the matching results.

[0020] It should be pointed out that in actual applications, there are various interferences, and the overlap between the first target three-dimensional point cloud set and the second target three-dimensional point cloud set is difficult to reach 100%. Therefore, the user can select a preset overlap according to needs, for example: setting the preset overlap to 90%, so that when the overlap is ≥90%, it is determined that the first ranging camera and the second ranging camera have detected the same target.

[0021] Further, in the method for generating a warning image based on a multi-focal-segment multi-camera group according to the present invention, in step S5, it is determined whether the target is complete in the image information of the second ranging camera, specifically: Obtaining an overlapping area of ​​images of the first ranging camera and the second ranging camera; When the target detected by the first ranging camera exceeds the upper edge, the left edge or the right edge of the picture overlap area, it is determined that the target is incomplete in the image information of the second ranging camera.

[0022] Furthermore, in the method for generating a warning image based on a multi-focal-segment multi-camera group described in the present invention, a feature point matching method is used to obtain the overlapping area of ​​the images of the first ranging camera and the second ranging camera.

[0023] Furthermore, the method for generating a warning image based on a multi-focal-segment multi-camera group according to the present invention also includes the steps of: setting up a test environment, placing targets of different sizes in the picture of the first ranging camera to detect the number of pixels of each target, and determining a preset pixel number threshold.

[0024] Correspondingly, another object of the present invention is to disclose a warning picture generation system based on a multi-focal-segment multi-eye camera group, which can easily execute the above-mentioned warning picture generation method of the present invention to ensure that the warning picture field is maximized under any circumstances, so that the user can see the ranging result more intuitively from the warning picture.

[0025] In the present invention, the warning image generation system based on a multi-focal-length multi-eye camera group specifically includes: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, each step of the warning image generation method based on a multi-focal-length multi-eye camera group as described above in the present invention is implemented.

[0026] The beneficial effects of the present invention are as follows: the present invention discloses a warning picture generation method based on a multi-focal-segment multi-eye camera group, the warning picture generation method can shoot a preset area through a first detection camera and a second ranging camera, and analyze the captured image information, so as to adopt different strategies to generate warning pictures according to different detection situations, the focal length of the first ranging camera is smaller than that of the second ranging camera, the field of view of the first ranging camera is larger, no matter what kind of detection situation is, the target warning information finally detected is superimposed on the image information of the first ranging camera to obtain the warning picture, thereby ensuring that the field of view of the warning picture is maximized in any situation, so that the user can see the ranging result more intuitively.

[0027] Correspondingly, the warning picture generation system based on a multi-focal-segment multi-eye camera group disclosed in the present invention can effectively execute the above-mentioned warning picture generation method of the present invention, and it also has the above-mentioned advantages and beneficial effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 The present invention is a flowchart of the steps of a method for generating a warning image based on a multi-focal-segment multi-eye camera group in one embodiment. DETAILED DESCRIPTION

[0029] In order to explain the technical content, achieved objectives and effects of the present invention in detail, the following is an explanation in combination with the implementation modes and the accompanying drawings.

[0030] See also Figure 1 As shown, the present invention specifically discloses a new warning image generation method based on a multi-focal-length multi-eye camera group; and, in this embodiment, the multi-focal-length multi-eye camera group specifically includes a first ranging camera and a second ranging camera, and the focal length of the first ranging camera is smaller than that of the second ranging camera; wherein, in this embodiment, the focal length of the first ranging camera is 8 mm, and the focal length of the second ranging camera is 25 mm.

[0031] The reason for setting up two ranging cameras with different focal lengths is that: when using a single high-resolution ranging camera for monitoring and ranging, there may be a problem of poor detection effect due to a single focal length.

[0032] Therefore, in this embodiment, the multi-focal-length multi-eye camera group simultaneously uses the first ranging camera and the second ranging camera to shoot and monitor the preset area, so as to improve its own detection capability, obtain a more comprehensive monitoring effect, and improve the accuracy and reliability of ranging.

[0033] To this end, based on the camera settings of the multi-focal-segment multi-eye camera group, when actually constructing the above-mentioned warning image generation method of the present invention, the following steps may be specifically included: S1: Use a multi-focal-length multi-eye camera group to shoot a preset area, and analyze the captured image information, so as to execute S2 or S3 when at least one ranging camera detects a target.

[0034] S2: When only the first ranging camera or the second ranging camera detects the target, warning information of the detected target is directly superimposed on the image information of the first ranging camera to obtain a warning image.

[0035] In the above technical solution of the present invention, before using the above-mentioned first ranging camera and second ranging camera to detect the target, after installing the first ranging camera and the second ranging camera, internal and external parameter calibration can be performed on each camera, and the posture relationship between the two ranging cameras can be obtained.

[0036] In addition, in this embodiment, the detected target warning information may specifically include: target segmentation result, ranging line and ranging value.

[0037] Based on this, for example, when the second ranging camera detects a target and obtains target warning information, the target warning information detected by the second ranging camera is superimposed on the image information of the first ranging camera, and a warning picture is obtained, which may specifically include the following contents: The position and posture relationship of each ranging camera has been obtained. First, according to the parameters of the second ranging camera and the detected target segmentation result, the target is projected into the three-dimensional space; and according to the parameters of the first ranging camera, the three-dimensional space target is projected onto the image information of the first ranging camera; then, according to the parameters of the second ranging camera, the two end points of the ranging line detected by the second ranging camera are projected into the three-dimensional space; and according to the parameters of the first ranging camera, the three-dimensional space end points are projected onto the image information of the first ranging camera; finally, the ranging value detected by the second ranging camera is directly superimposed on the image information of the first ranging camera, and the final warning picture is obtained.

[0038] S3: When the first ranging camera and the second ranging camera detect the same target, the number of pixels of the target in the image information of the first ranging camera is obtained, and S4 or S5 is executed.

[0039] It should be pointed out that in actual applications, when the first detection camera and the second detection camera are used to monitor the preset area at the same time, in some specific circumstances, the first detection camera and the second detection camera will inevitably detect the target at the same time. At this time, the targets detected in the first detection camera and the second detection camera can be matched according to the coordinate mapping relationship of the pixel points on the two cameras, so as to execute the appropriate warning image generation strategy according to the matching results.

[0040] To this end, in the present invention, in the above step S3, the above judgment process specifically includes the steps of: The position and posture relationship of each ranging camera has been obtained. First, according to the parameters of the first ranging camera and the detected target segmentation result, the target is projected into the three-dimensional space to obtain a first target three-dimensional point cloud set; then, according to the parameters of the first ranging camera and the detected target segmentation result, the target is projected into the three-dimensional space to obtain a second target three-dimensional point cloud set; finally, the overlap between the first target three-dimensional point cloud set and the second target three-dimensional point cloud set is compared. When the overlap reaches a preset overlap, it is determined that the first ranging camera and the second ranging camera have detected the same target.

[0041] It should be pointed out that in actual applications, there are various interferences, and the overlap between the first target three-dimensional point cloud set and the second target three-dimensional point cloud set is difficult to reach 100%. Therefore, the user can select a preset overlap according to needs, for example: setting the preset overlap to 90%, so that when the overlap is ≥90%, it is determined that the first ranging camera and the second ranging camera have detected the same target.

[0042] S4: if the number of pixels is greater than or equal to a preset pixel number threshold M, the target warning information detected by the first ranging camera is directly superimposed on the image information of the first ranging camera to obtain a warning image.

[0043] S5: When the number of pixels is less than the preset pixel number threshold M, it is determined whether the target is complete in the image information of the second ranging camera; if complete, the target warning information detected by the second ranging camera is superimposed on the image information of the first ranging camera to obtain a warning picture; if incomplete, the target warning information detected by the first ranging camera is superimposed on the image information of the first ranging camera to obtain a warning picture.

[0044] In the above step S4 and step S5 of the present invention, in order to ensure that the target can be well displayed in the generated warning image; in practical application, the warning image can also be flexibly generated through the pixel points of the target, specifically: a test environment is pre-built, and targets of different sizes are placed in the picture of the first ranging camera to detect the number of pixels of each target, and a preset pixel number threshold M is determined through actual measurement; that is, when the pixel number of the target in the first ranging camera is ≥ the preset pixel number threshold M, the measurement accuracy of the first ranging camera for the target is good, and the generation strategy of the above step S4 is specifically executed; when the pixel number of the target in the first ranging camera is < the preset pixel number threshold M, the measurement accuracy of the first ranging camera for the target is poor, and the generation strategy of step S5 is specifically executed.

[0045] In addition, it should be noted that, when the above step S5 is actually executed, in order to further optimize the generation effect of the final warning picture, it is also necessary to determine whether the target is complete in the image information of the second ranging camera, which may specifically include the following steps: A feature point matching method or other method is used to obtain the overlapping area of ​​the images of the first ranging camera and the second ranging camera; and when the target detected by the first ranging camera exceeds the upper edge, the left edge or the right edge of the overlapping area of ​​the image, it is determined that the target is incomplete in the image information of the second ranging camera; in other cases, it is determined that the target is complete in the image information of the second ranging camera.

[0046] The reason for adopting the above setting form is that the warning image generation method of the present invention is usually applied in the field of power transmission, and specifically tests the distance from the target to the transmission line to ensure the safety of power transmission; therefore, if the target detected by the above-mentioned ranging camera exceeds the lower edge of the above-mentioned overlapping area of ​​the picture, it does not affect the calculation of this distance, while the upper half of the target has a greater impact on the calculation of this distance.

[0047] Of course, in the above step S5 of the present invention, when it is determined that the above target is complete in the image information of the second ranging camera, the target warning information detected by the second ranging camera will be superimposed on the image information of the first ranging camera to obtain a warning picture; the specific superposition process can be referred to the commentary of the above step S2, and will not be repeated here.

[0048] In addition, it should be noted that in the present invention, when it is actually applied, considering that the first rangefinder camera and the second rangefinder camera are combined in a single device for snapshot ranging, there are still certain usage defects; for this reason, when it is actually used, a multi-focal-segment multi-eye camera group can also be set up to include a main device and a sub-device, and the main device and the sub-device are respectively set at two different positions of the transmission tower, and the main device and the sub-device are controlled to include the first rangefinder camera and the second rangefinder camera, so as to simultaneously use the first rangefinder camera and the second rangefinder camera in the main device and the sub-device to shoot the preset area, and respectively perform detection and ranging. Among them, when assembling the above-mentioned main device and sub-device, the optical axes of the first rangefinder camera and the second rangefinder camera in the same device are parallel to ensure the accuracy of subsequent detection.

[0049] At the same time, after installing the above-mentioned main camera and auxiliary camera, the internal and external parameters of the four ranging cameras in the main camera and auxiliary camera can be calibrated, and the position and posture relationship between the four ranging cameras can be obtained, so as to facilitate the subsequent establishment of a mapping relationship and obtain an alarm picture.

[0050] Of course, in order to avoid the problem of user confusion caused by uploading multiple alarm pictures, in actual application, no matter what the situation is, which camera detects the target alarm information, it is finally superimposed on the image information of the first ranging camera of the main device to obtain an alarm picture, which can effectively facilitate power grid workers to obtain the ranging results intuitively and accurately, and judge whether the distance between the target and the transmission cable is within a safe distance.

[0051] Correspondingly, in the present invention, a warning picture generation system based on a multi-focal-segment multi-eye camera group is also disclosed. The warning picture generation system can simply execute the above-mentioned warning picture generation method of the present invention to flexibly generate warning pictures based on the snapshot ranging conditions of the multi-focal-segment multi-eye camera group, which specifically includes: A memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, each step of the method for generating a warning image based on a multi-focal-segment multi-eye camera group as described above in the present invention is implemented, that is, the above steps are executed.

[0052] From the above description, it can be seen that the warning picture generation method and system based on the multi-focal-length multi-eye camera group designed by the present invention can solve two problems that may exist in the existing scheme: uploading a long-focus warning picture may result in the situation that the wire cannot be seen and the key scene cannot be seen completely; if the target appears in multiple ranging cameras at the same time, the problem of two ranging results for the same target may occur. When the warning picture generation method is actually applied to the multi-focal-length multi-eye camera group, no matter which ranging camera detects the target and obtains the target warning information, it is finally superimposed on the image information collected by the first ranging camera with a short focal length to generate a warning picture. Uploading the short-focal-length ranging camera screen as a warning picture can allow users to have a more intuitive feeling of the target position, ranging results and the overall scene. It can ensure that the warning picture field is maximized in any case, so that users can see the ranging results more intuitively. It has high reliability and is easy to implement, and has good promotion prospects and application value.

[0053] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent transformations made using the contents of the present invention's specification and drawings, or directly or indirectly applied in related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for generating a warning image based on a multi-focal-segment multi-eye camera group, wherein the multi-focal-segment multi-eye camera group includes a first ranging camera and a second ranging camera, wherein the focal length of the first ranging camera is smaller than that of the second ranging camera, and wherein: Includes steps: S1: photographing a preset area using a multi-focal-length multi-eye camera group, and analyzing the captured image information, so as to execute S2 or S3 when at least one ranging camera detects a target; S2: Only the first ranging camera or the second ranging camera detects the target, and directly superimposes the detected target warning information on the image information of the first ranging camera to obtain a warning picture; S3: If the first ranging camera and the second ranging camera detect the same target, the number of pixels of the target in the image information of the first ranging camera is obtained, and S4 or S5 is executed; S4: the number of pixels is ≥ the preset pixel number threshold, directly superimposing the target warning information detected by the first ranging camera on the image information of the first ranging camera to obtain a warning picture; S5: If the number of pixels is less than the preset pixel number threshold, it is determined whether the target is complete in the image information of the second ranging camera; if it is complete, the target warning information detected by the second ranging camera is superimposed on the image information of the first ranging camera to obtain a warning picture; if it is incomplete, the target warning information detected by the first ranging camera is superimposed on the image information of the first ranging camera to obtain a warning picture.

2. The method for generating warning images based on a multi-focal-segment multi-camera group according to claim 1, characterized in that: The multi-focal-length multi-eye camera group includes a main device and a sub-device, and the main device and the sub-device both include a first ranging camera and a second ranging camera. In step S1, the first ranging camera and the second ranging camera in the main device and the sub-device are used simultaneously to shoot a preset area.

3. The method for generating warning images based on a multi-focal-segment multi-camera group according to claim 2, characterized in that: In steps S2, S4 and S5, the detected target warning information is superimposed on the image information of the first ranging camera of the master device to obtain a warning picture.

4. The method for generating warning images based on a multi-focal-segment multi-camera group according to claim 1, characterized in that: The target warning information includes: target segmentation result, ranging line and ranging value.

5. The method for generating warning images based on a multi-focal-segment multi-camera group according to claim 4, characterized in that: In step S2 or step S5, the target warning information detected by the second ranging camera is superimposed on the image information of the first ranging camera, specifically: According to the parameters of the second ranging camera and the detected target segmentation result, the target is projected into the three-dimensional space; and according to the parameters of the first ranging camera, the three-dimensional space target is projected onto the image information of the first ranging camera; According to the parameters of the second ranging camera, the two end points of the ranging line detected by the second ranging camera are projected into the three-dimensional space; and according to the parameters of the first ranging camera, the three-dimensional space end points are projected onto the image information of the first ranging camera; The distance measurement value detected by the second distance measurement camera is directly superimposed on the image information of the first distance measurement camera.

6. The method for generating warning images based on a multi-focal-segment multi-camera group according to claim 4, characterized in that: In step S3, it is determined whether the first ranging camera and the second ranging camera detect the same target, specifically: According to the parameters of the first ranging camera and the detected target segmentation result, the target is projected into a three-dimensional space to obtain a first target three-dimensional point cloud set; According to the parameters of the first ranging camera and the detected target segmentation result, the target is projected into a three-dimensional space to obtain a second target three-dimensional point cloud set; The overlap between the first target three-dimensional point cloud set and the second target three-dimensional point cloud set is compared, and when the overlap reaches a preset overlap, it is determined that the first ranging camera and the second ranging camera have detected the same target.

7. The method for generating warning images based on a multi-focal-segment multi-camera group according to claim 1, characterized in that: In step S5, it is determined whether the target is complete in the image information of the second ranging camera, specifically: Obtaining an overlapping area of ​​images of the first ranging camera and the second ranging camera; When the target detected by the first ranging camera exceeds the upper edge, the left edge or the right edge of the picture overlap area, it is determined that the target is incomplete in the image information of the second ranging camera.

8. The method for generating warning images based on a multi-focal-segment multi-camera group according to claim 7, characterized in that: The overlapping area of ​​the images of the first ranging camera and the second ranging camera is obtained by using a feature point matching method.

9. The method for generating warning images based on a multi-focal-segment multi-camera group according to claim 1, characterized in that: The method also includes the steps of: setting up a test environment, placing targets of different sizes in the picture of the first ranging camera to detect the number of pixels of each target, and determining a preset pixel number threshold.

10. A warning picture generation system based on a multi-focal-segment multi-eye camera group, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, each step of the method for generating a warning image based on a multi-focal-segment multi-eye camera group as described in any one of claims 1 to 9 is implemented.

Images