Camera pose calculation method and device based on point cloud picture and computer readable storage medium

Abstract

The invention belongs to the technical field of power transmission line image analysis, and more specifically relates to a camera pose calculation method and device based on a point cloud picture, and a computer readable storage medium. The method comprises the following steps: acquiring point cloud data, namely power transmission scene point cloud, of a scene around a power transmission line, and shooting a power transmission scene image through visual monitoring and shooting equipment; obtaining a tower area in the power transmission scene image through a detection algorithm; taking the tower area as a center, and obtaining an image area coinciding with the power transmission scene point cloud in the power transmission scene image; removing a tower point cloud where the camera is located in the point cloud data, and obtaining an overlapped point cloud region corresponding to the overlapped image region; matching between the image and the point cloud is completed; and obtaining a camera pose. According to the invention, the problems that there is no automatic mode to obtain the pose of the camera at present and the pose of the camera is difficult to obtain by a power transmission scene visualization device at present are solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of power transmission line image analysis, and more specifically, relates to a camera pose calculation method and device based on point cloud images and a computer-readable storage medium. Background Art

[0002] With the rapid development of the national economy, the number of power transmission lines, as an important part of the power system, is increasing. Its reliable and safe operation is directly related to the reliability of power supply of the power grid. However, in the actual operation of power transmission lines, there are various hidden dangers. Among them, mechanical hidden dangers, trees, and conductor changes in the line corridor are great safety hazards of transmission lines, which also seriously affect the safe operation of transmission lines. How to timely discover, handle and prevent tree barrier hidden dangers and ensure the safety of power transmission has become an important issue of concern to transmission operation and maintenance personnel.

[0003] Chinese invention patent CN115641373A discloses an interactive 3D ranging algorithm integrating point cloud and image, including S1, wire data segmentation and annotation, 2D-3D point pair data matching and annotation, and obtaining annotated scene data; S2, initial pose calculation; S3, judging whether the initial pose meets the requirements, that is, whether the 2D-3D point pairs in step S1 have overlapped, if so, proceed to S4; otherwise, proceed to S4 after manually aligning the pose; S4, point cloud pixel preprocessing; S5, 3D ranging.

[0004] With the development of modern science and technology, more and more visual monitoring equipment is installed on transmission line towers to replace manual inspections of transmission lines. If the capabilities of visual monitoring equipment for transmission lines need to be better utilized, it is essential to obtain the camera pose. After obtaining the camera pose, simple multi-source data fusion of related transmission points and automated protection zone delineation can be achieved, providing basic data support for the delineation of hidden danger levels in transmission scenes based on visual monitoring equipment, thus ensuring the safety of transmission lines. However, there is currently no automated way to obtain the camera pose, and it is difficult for current transmission scene visualization equipment to obtain the camera pose. Summary of the invention

[0005] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a camera pose calculation method based on point cloud images, which uses the point cloud around the transmission line obtained by laser radar scanning, RGBD camera scanning, visual reconstruction and other methods and the pictures taken by the visualization monitoring equipment as the data source, obtains the position of the tower in the image through the tower target detection algorithm, and then obtains the tower area, and then obtains the image area that can overlap with the point cloud with the tower area as the center, and at the same time, removes the tower where the camera is located in the transmission line point cloud through the coordinates of the tower position where the camera is located, and obtains the point cloud of the overlapping area with the image, and then uses the CorrI2P and other algorithms to realize the overlapping image area and the coordinates or numbers of the same points in the point cloud of the overlapping area with the image, and finally uses the camera pose estimation algorithm to obtain the camera pose of the visualization monitoring equipment of the transmission scene.

[0006] The technical solution adopted by the present invention to achieve the above-mentioned purpose is:

[0007] A camera pose calculation method based on a point cloud image, characterized in that it comprises the following steps:

[0008] S1. Obtain point cloud data of the scene around the transmission line, i.e., the transmission scene point cloud. At the same time, take images of the transmission scene through a visual monitoring device;

[0009] S2, obtaining the tower area tower_box in the power transmission scene image through the detection algorithm;

[0010] S3, taking the tower area as the center, obtaining the image area img_cut in the power transmission scene image that coincides with the power transmission scene point cloud;

[0011] S4, according to the position coordinates of the pole tower where the camera is located, removing the pole tower point cloud where the camera is located in the point cloud data, and obtaining the overlapping point cloud area corresponding to the overlapping image area;

[0012] S5. Use the CorrI2P algorithm to obtain the coordinates or numbers of the same points in the overlapping image area and the overlapping point cloud area, and complete the matching between the image and the point cloud;

[0013] The coordinates or numbers of the same points in the overlapping image area and the overlapping point cloud area are used to describe the features of the same position of the same scene. At the same time, the coordinate points of the point cloud points and the image of the same scene are converted to each other through the camera imaging principle;

[0014] The CorrI2P algorithm is as follows: the input is the image area with overlapping point clouds and the point cloud of the image overlapping area, and the output is the neural network algorithm model of the coordinates or numbers of the same points in the overlapping image area and the point cloud of the overlapping image area.

[0015] S6. Based on the matching between the image and the point cloud, a camera pose estimation algorithm is used to obtain the camera pose of the power transmission scene visualization monitoring device.

[0016] The camera pose estimation algorithm takes as input the overlapping image area and the coordinates of the same points in the point cloud of the overlapping area and the camera intrinsic parameter matrix, and outputs an algorithm model of the camera pose of the power transmission scene visualization monitoring device.

[0017] Furthermore, the S2 specifically includes:

[0018] S21, obtaining all tower hidden danger frames tower_bbox in the image through a tower target detection algorithm, and then obtaining a segmentation result seg, i.e., a tower in the image, through an image segmentation algorithm. The tower in the image refers to the first tower in the direction of the camera lens where the camera is installed;

[0019] The segmentation result seg is a matrix of 1944*2592*1, each element of which is a value of -1 or -2, where -1 represents non-tower pixels and -2 represents tower pixels.

[0020] S22, calculating the number of poles and towers in each hidden danger frame based on the segmentation results;

[0021] First, obtain the segmentation result information of the area of ​​the ith tower hidden danger box:

[0022] seg_copy[i]=seg[tower_bbox[i][0]:tower_bbox[i][2],tower_bbox[i][1]:tower_bbox[i][3]] (1);

[0023] In formula (1), seg_copy[i] is the segmentation result information of the hidden danger box of the i-th tower, tower_bbox[i][0] is the x-coordinate of the upper left corner of the hidden danger box of the i-th tower, tower_bbox[i][2] is the x-coordinate of the lower right corner of the hidden danger box of the i-th tower, tower_bbox[i][1] is the y-coordinate of the hidden danger box of the i-th tower, and tower_bbox[i][3] is the y-coordinate of the hidden danger box of the i-th tower.

[0024] Then calculate the number of poles and towers for the segmented pixels of all the detected tower hidden danger boxes:

[0025] tower_bbox_seg_mount[i]=mount(seg_copy[i]==-2) (2);

[0026] In formula (2), tower_bbox_seg_mount[i] is the number of pixels in the hidden danger box of the i-th tower that are towers, and mount is the operation of taking the number of elements in seg_copy[i] whose value is equal to -2.

[0027] S23, selecting the tower hidden danger box with the largest number of tower pixels as the tower area tower_box;

[0028] tower_box=tower_bbox[argmax(tower_bbox_seg_mount)] (3);

[0029] In formula (3), argmax is an operation to obtain the maximum numerical subscript of the elements in the array tower_bbox_seg_mount in the brackets.

[0030] Furthermore, the overlapping image area img_cut is a matrix of (tower_box[3]-tower_box[1]+500)*(tower_box[2]-tower_box[4]+500)*3;

[0031] img_cut=[tower_box[0]-250,tower_box[1]-250,tower_box[2]+250,tower_box[3]+250] (4);

[0032] In formula (4), tower_box[0] is the x-coordinate of the upper left corner of the tower area, tower_box[1] is the y-coordinate of the upper left corner of the tower area, tower_box[2] is the x-coordinate of the lower right corner of the tower area, and tower_box[3] is the y-coordinate of the lower right corner of the tower area.

[0033] Furthermore, the S4 specifically includes:

[0034] S41, obtaining the position coordinates of the tower in the power transmission scene image;

[0035] S42, calculating the xy plane Euclidean distance OU1 between the position coordinates of the tower in the power transmission scene image and the position coordinates of the tower where the camera is located;

[0036] S43, taking the coordinates of the tower where the camera is located as the center, delete all points in the transmission channel tower point cloud whose xy plane Euclidean distance is less than 0.5 times OU1, and the corresponding overlapping point cloud areas of the overlapping image areas.

[0037] In another aspect of the present invention, a device for implementing a camera pose calculation method based on a point cloud image is provided, the device comprising:

[0038] at least one processor; and

[0039] A memory storing instructions, which, when executed by the at least one processor, enables the at least one processor to execute the camera pose calculation method based on the point cloud image as described above.

[0040] In another aspect of the present invention, a machine-readable storage medium is provided, which stores executable instructions, and when the instructions are executed, the machine executes the camera pose calculation method based on point cloud images as described above.

[0041] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0042] The present invention discloses a camera pose calculation method and device based on point cloud images, and a computer-readable storage medium. By processing point clouds and images, the point cloud image matching area is determined, and the camera pose acquisition of power transmission scene visualization monitoring equipment is realized, which provides reliable data support for rapid multi-source data fusion and effectively solves the problem that it is difficult to obtain camera pose for current power transmission scene visualization equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a schematic flow chart of the method of the present invention.

[0044] Figure 2 This is a schematic diagram of the image visualization effect of the power transmission scene on the tower where the camera is located in Example 1.

[0045] Figure 3 It is a schematic diagram of the visualization effect of the image area where the tower area and the point cloud overlap in Example 1.

[0046] Figure 4 It is a schematic diagram of the visualization effect of the overlapping point cloud area after removing the point cloud of the tower where the camera is located in Example 1. DETAILED DESCRIPTION

[0047] The present invention will be further described below in conjunction with the embodiments.

[0048] Example 1

[0049] This embodiment provides a camera pose calculation method based on point cloud images, such as Figure 1 As shown in the figure, the data used are the image of a certain point on a transmission line and point cloud data, including:

[0050] S1. Obtain the point cloud and image of the power transmission scene:

[0051] Obtain point cloud data of the scene around the transmission line, i.e., the transmission scene point cloud. At the same time, use visual monitoring equipment to capture images of the transmission scene.

[0052] Preferably, point cloud data is obtained by using laser radar scanning, RGBD camera scanning, visual reconstruction and other methods; specifically, the point cloud data of the scene around the transmission line refers to the position information of the scene around the transmission line in space obtained by means of laser radar and other means, mainly represented by a group of n*3 arrays, and the point cloud data

[0053] At the same time, the power transmission scene image is captured by a visual monitoring device for the power transmission scene, such as Figure 2 As shown, the image matrix img is a 1944*2592*3 matrix, and its visualization effect is as follows Figure 3 As shown, the coordinates of the tower where the camera is located are tower_local = [-3.20, 0, 56.2], and the camera internal parameter matrix of the visual monitoring device is

[0054] S2. Get the tower area in the image:

[0055] The tower area tower_box in the power transmission scene image is obtained through the detection algorithm, such as Figure 2 As shown in the blue box, preferably, the detection algorithm is implemented based on RCNN and YOLOv5 algorithms.

[0056] Preferably, S2 in this embodiment specifically includes:

[0057] S21. Obtain all tower hidden danger frames tower_bbox in the image through a tower target detection algorithm. Specifically, in this embodiment, two tower hidden danger frames tower_bbox=[[1286,764,1385,997],[830,263,920,440]] are detected.

[0058] Then, the segmentation result seg is obtained by an image segmentation algorithm, that is, the pole tower in the image, where the pole tower in the image refers to the first pole tower in the direction of the camera lens in which the camera is installed;

[0059] Specifically, the segmentation result seg is obtained through an image segmentation algorithm, where seg is a matrix of 1944*2592*1, each element is a value of -1 or -2, -1 represents non-tower pixels, and -2 represents tower pixels. Preferably, the segmentation algorithm is a neural network algorithm such as Pointrend.

[0060] S22, calculating the number of pixels in each hidden danger frame that are pole towers based on the pole tower segmentation result;

[0061] Since most of the existing segmentation results are relatively poor, it is impossible to directly obtain the complete tower area, so it is necessary to combine the target detection algorithm to obtain the tower area;

[0062] First, obtain the segmentation result information of the area of ​​the ith tower hidden danger box:

[0063] Let i take the value from 0 to 1, then

[0064] seg_copy[i]=seg[tower_bbox[i][0]:tower_bbox[i][2],tower_bbox[i][1]:tower_bbox[i][3]] (1);

[0065] Among them, seg_copy[i] is the segmentation result information of the hidden danger box of the i-th tower, tower_bbox[i][0] is the x-coordinate of the upper left corner of the hidden danger box of the i-th tower, tower_bbox[i][2] is the x-coordinate of the lower right corner of the hidden danger box of the i-th tower, tower_bbox[i][1] is the y-coordinate of the hidden danger box of the i-th tower, and tower_bbox[i][3] is the y-coordinate of the hidden danger box of the i-th tower.

[0066] For example: when i == 0, seg_copy[0] = seg[[1286:1385,764:997]];

[0067] When i==1, seg_copy[1]=seg[[830,263,920,440]].

[0068] Then calculate the number of poles and towers for the segmented pixels of all the detected tower hidden danger boxes:

[0069] tower_bbox_seg_mount[i]=mount(seg_copy[i]==-2) (2);

[0070] In formula (2), tower_bbox_seg_mount[i] is the number of pixels in the hidden danger box of the i-th tower that are towers; mount is the operation of taking the number of elements in seg_copy[i] whose value is equal to -2.

[0071] S23, selecting the tower hidden danger box with the largest number of tower pixels as the tower area tower_box;

[0072] tower_box=tower_bbox[argmax(tower_bbox_seg_mount)] (3);

[0073] In formula (3), argmax is an operation to obtain the maximum numerical subscript of the elements in the array tower_bbox_seg_mount in the brackets.

[0074] S3. Get the overlapping image area:

[0075] Taking the tower area as the center, obtain the image area img_cut in the power transmission scene image that coincides with the power transmission scene point cloud;

[0076] The power transmission scene image and the power transmission scene point cloud describe the same thing, and the point cloud is a spatial description of the thing in the image area.

[0077] Generally speaking, the transmission channel area captured by the image is wider than the area scanned by the LiDAR. This results in low algorithm accuracy when aligning point cloud data due to the wide image capture area.

[0078] Specifically, the overlapping area is as follows: Figure 2 As shown in the white box, the overlapping image area img_cut is a matrix of (tower_box[3]-tower_box[1]+500)*(tower_box[2]-tower_box[0]+500)*3, that is, Figure 2 Coordinates of the middle white box;

[0079] img_cut=[tower_box[0]-250,tower_box[1]-250,tower_box[2]+250,tower_box[3]+250] (4);

[0080] In formula (4), tower_box[0] is the x-coordinate of the upper left corner of the tower area, tower_box[1] is the y-coordinate of the upper left corner of the tower area, tower_box[2] is the x-coordinate of the lower right corner of the tower area, and tower_box[3] is the y-coordinate of the lower right corner of the tower area.

[0081] S4. Get overlapping point cloud area:

[0082] According to the coordinates of the tower where the camera is located, the tower point cloud where the camera is located is removed from the point cloud data to obtain the overlapping point cloud area corresponding to the overlapping image area. The obtained point cloud is as follows: Figure 4 shown.

[0083] The coordinates are based on the point cloud data coordinate system as the reference coordinate system, and the coordinate position is the center point of the contact plane between the tower and the ground;

[0084] The point cloud in the image overlapping area is the remaining point cloud of the transmission line after removing the point cloud of the tower where the camera is located, including but not limited to common scenes in power transmission scenes such as conductors, opposite towers, ground, vegetation, trees, and buildings.

[0085] Furthermore, the S4 specifically includes:

[0086] S41, obtaining the position coordinates of the tower in the power transmission scene image (the position coordinates are unique);

[0087] S42, calculating the xy plane Euclidean distance OU1 between the position coordinates of the tower in the power transmission scene image and the position coordinates of the tower where the camera is located;

[0088] S43, taking the coordinates of the tower where the camera is located as the center, delete all points in the transmission channel tower point cloud whose xy plane Euclidean distance is less than 0.5 times OU1, and the corresponding overlapping point cloud areas of the overlapping image areas.

[0089] S5. Matching between image and point cloud:

[0090] Use the CorrI2P algorithm to obtain the coordinates or numbers of the same points in the overlapping image area and the overlapping point cloud area, and complete the matching between the image and the point cloud;

[0091] Preferably, the obtained overlapping image area and the coordinates of the same points in the point cloud of the image overlapping area, wherein the image coordinates matching the point cloud are im=[[1583,977],…,[1582,979]], and the corresponding point cloud coordinates are pc_l=[[169.86,388.36,-14.77],…,[69.83,388.46,-14.99]], and the group of data with the same subscripts as im and pc_l are the corresponding same points, such as im[0]=[1583,977], pc_1[0]=[169.86,388.36,-14.77], where 0 is the subscript, which is also the position of the element in the array in the array, starting from 0.

[0092] S6. Get the camera pose:

[0093] According to the matching between the image and the point cloud, the camera pose estimation algorithm is used to obtain the camera pose of the power transmission scene visualization monitoring equipment.

[0094] The camera pose estimation algorithm preferably uses the slovePnPRanscen function in the opencv2 package in Python as an example to obtain the camera pose of the power transmission scene visualization monitoring device, and finally obtains the camera rotation matrix R=[[0.99862897,-0.00690724,0.05188787],[-0.05227776,-0.18205254,0.98189813],[0.00266411,-0.98326457,-0.18216404]], and the camera translation matrix T=[-3.19,0,56.0].

[0095] Example 2

[0096] This embodiment provides a device for implementing a camera pose calculation method based on a point cloud image, the device comprising:

[0097] at least one processor; and

[0098] A memory storing instructions, which, when executed by the at least one processor, enables the at least one processor to execute the camera pose calculation method based on the point cloud image as described above.

[0099] In this embodiment, the device includes but is not limited to: personal computers, server computers, workstations, desktop computers, laptop computers, notebook computers, mobile computing devices, smart phones, tablet computers, cellular phones, personal digital assistants (PDAs), handheld devices, messaging devices, wearable computing devices, consumer electronic devices, etc.

[0100] Example 3

[0101] This embodiment also provides a computer-readable storage medium storing executable instructions, which, when executed, enable the machine to perform the camera pose calculation method based on point cloud images as described above.

[0102] Specifically, a system or device equipped with a readable storage medium can be provided, on which software program codes that implement the functions of any of the above-mentioned embodiments are stored, and a computer or processor of the system or device can read and execute instructions stored in the readable storage medium.

[0103] In this case, the program code itself read from the machine-readable medium can realize the function of any one of the above embodiments, and thus the machine-readable code and the machine-readable storage medium storing the machine-readable code constitute part of this specification.

[0104] Examples of readable storage media include floppy disks, hard disks, magneto-optical disks, optical disks (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD-RW), magnetic tapes, non-volatile memory cards, and ROMs. Alternatively, the program code may be downloaded from a server computer or a cloud via a communication network.

[0105] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0106] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0107] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0108] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0109] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the technical solution of the present invention, and are not intended to limit the specific implementation methods of the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the claims of the present invention shall be included in the protection scope of the claims of the present invention.

Claims

1. A camera pose calculation method based on point cloud images, characterized in that: include: S1. Obtain point cloud data of the scene around the transmission line, i.e., the transmission scene point cloud. At the same time, take images of the transmission scene through a visual monitoring device; S2, obtaining the tower area tower_box in the power transmission scene image through the detection algorithm; S3, taking the tower area as the center, obtaining the image area img_cut in the power transmission scene image that coincides with the power transmission scene point cloud; S4, according to the position coordinates of the pole tower where the camera is located, removing the pole tower point cloud where the camera is located in the point cloud data, and obtaining the overlapping point cloud area corresponding to the overlapping image area; S5. Use the CorrI2P algorithm to obtain the coordinates or numbers of the same points in the overlapping image area and the overlapping point cloud area, and complete the matching between the image and the point cloud; S6. Based on the matching between the image and the point cloud, a camera pose estimation algorithm is used to obtain the camera pose of the power transmission scene visualization monitoring device.

2. The camera pose calculation method based on point cloud images according to claim 1, characterized in that: The S2 specifically includes: S21, obtaining all tower hidden danger frames tower_bbox in the image through a tower target detection algorithm, and then obtaining a segmentation result seg, i.e., a tower in the image, through an image segmentation algorithm. The tower in the image refers to the first tower in the direction of the camera lens where the camera is installed; The segmentation result seg is a matrix of 1944*2592*1, each element of which is a value of -1 or -2, where -1 represents non-tower pixels and -2 represents tower pixels; S22, calculating the number of poles and towers in each hidden danger frame based on the segmentation results; First, obtain the segmentation result information of the area of ​​the ith tower hidden danger box: seg_copy[i]=seg[tower_bbox[i][0]:tower_bbox[i][2],tower_bbox[i][1]:tower_bbox[i][3]] (1); In formula (1), seg_copy[i] is the segmentation result information of the hidden danger box of the i-th tower, tower_bbox[i][0] is the x-coordinate of the upper left corner of the hidden danger box of the i-th tower, tower_bbox[i][2] is the x-coordinate of the lower right corner of the hidden danger box of the i-th tower, tower_bbox[i][1] is the y-coordinate of the hidden danger box of the i-th tower, and tower_bbox[i][3] is the y-coordinate of the hidden danger box of the i-th tower; Then calculate the number of poles and towers for the segmented pixels of all the detected tower hidden danger boxes: tower_bbox_seg_mount[i]=mount(seg_copy[i]==-2) (2); In formula (2), tower_bbox_seg_mount[i] is the number of towers whose pixels in the hidden danger box of the i-th tower are calculated; mount is the operation of taking the number of elements in seg_copy[i] whose value is equal to -2; S23, selecting the tower hidden danger box with the largest number of tower pixels as the tower area tower_box; tower_box=tower_bbox[argmax(tower_bbox_seg_mount)] (3); In formula (3), argmax is an operation to obtain the maximum numerical subscript of the elements in the array tower_bbox_seg_mount in the brackets.

3. The camera pose calculation method based on point cloud images according to claim 2, characterized in that: The overlapping image area img_cut is a matrix of (tower_box[3]-tower_box[1]+500)*(tower_box[2]-tower_box[4]+500)*3; img_cut=[tower_box[0]-250,tower_box[1]-250,tower_box[2]+250,tower_box[3]+250] (4); In formula (4), tower_box[0] is the x-coordinate of the upper left corner of the tower area, tower_box[1] is the y-coordinate of the upper left corner of the tower area, tower_box[2] is the x-coordinate of the lower right corner of the tower area, and tower_box[3] is the y-coordinate of the lower right corner of the tower area.

4. The camera pose calculation method based on point cloud images according to claim 3, characterized in that: The S4 specifically includes: S41, obtaining the position coordinates of the tower in the power transmission scene image; S42, calculating the xy plane Euclidean distance OU1 between the position coordinates of the tower in the power transmission scene image and the position coordinates of the tower where the camera is located; S43, taking the coordinates of the tower where the camera is located as the center, delete all points in the transmission channel tower point cloud whose xy plane Euclidean distance is less than 0.5 times OU1, and the corresponding overlapping point cloud areas of the overlapping image areas.

5. The camera pose calculation method based on point cloud images according to claim 1, characterized in that: The coordinates or numbers of the same points in the overlapping image area and the overlapping point cloud area are used to describe the characteristics of the same position of the same scene. At the same time, the point cloud points and the coordinate points of the image of the same scene are converted into each other through the camera imaging principle.

6. A camera pose calculation device based on point cloud images, characterized in that: The device comprises: processor; a memory having stored thereon a computer program executable on the processor; Wherein, when the computer program is executed by the processor, the steps of a camera pose calculation method based on point cloud images as described in any one of claims 1 to 5 are implemented.

7. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images