CN118429342B - Calibration image screening method, calibration method and related equipment - Google Patents

Abstract

The invention relates to the technical field of camera calibration, and particularly discloses a calibration image screening method, a calibration method and related equipment, wherein the calibration image screening method comprises the following steps: acquiring a shooting video containing a checkerboard calibration plate; extracting a captured image set from a captured video; candidate angular points in each image in the shot image set are obtained; screening and obtaining a first image set according to the angular point distribution rule of the checkerboard calibration plate, and determining reference angular points in each image in the first image set from candidate angular points, wherein the images in the first image set are images containing the checkerboard calibration plate in the shot image set; according to the calibration image screening method, incomplete image screening treatment, fuzzy image screening treatment and repeated image screening treatment are sequentially carried out on the shot image set, so that a calibration image set which finally contains complete clear and unrepeated effective pictures is obtained, the images in the calibration image set are ensured to be suitable for internal reference calibration of a camera, and the calibration precision and efficiency can be effectively improved.

Description

Calibration image screening method, calibration method and related equipment

Technical Field

The present application relates to the field of camera calibration technology, and in particular to a calibration image screening method, a calibration method and related equipment.

Background Art

Camera calibration is used to establish the correspondence between three-dimensional space points and image points in image measurement.

In the related technology, Zhang Zhengyou's calibration method uses an algorithm to calculate the camera parameters by shooting a chessboard image, and determines the conversion relationship between three-dimensional space points and pixel points and the distortion coefficient. When shooting, the chessboard image needs to be stabilized, clear pictures need to be taken from multiple angles, and then the camera internal parameters are calculated through the algorithm.

For the calibration process of multiple cameras, it is necessary to manually screen clear and effective pictures frame by frame for calibration, which leads to the problem of serious time consumption and low calibration efficiency for the entire camera calibration process.

There is currently no effective technical solution to the above problems.

Summary of the invention

The purpose of this application is to provide a calibration image screening method, a calibration method and related equipment to achieve automatic acquisition of clear and effective pictures to improve the efficiency of camera calibration.

In a first aspect, the present application provides a calibration image screening method for screening images that can be used for camera calibration from video data, wherein the camera calibration is performed based on a checkerboard calibration plate, and the calibration image screening method comprises the following steps:

S1, obtaining a captured video containing a checkerboard calibration plate;

S2, extracting a set of captured images from the captured video;

S3, obtaining candidate corner points in each image in the captured image set;

S4, selecting and acquiring a first image set according to the distribution rule of the corner points of the checkerboard calibration plate and determining the reference corner points in each image in the first image set from the candidate corner points, wherein the images in the first image set are the images in the captured image set that include the checkerboard calibration plate;

S5, removing blurred images in the first image set according to the image variance of the area around the reference corner point to select a second image set;

S6. Remove duplicate images in the second image set according to the positional relationship of the reference corner points of the edge to screen out a calibration image set.

The calibration image screening method of the present application sequentially performs incomplete image screening processing, blurred image screening processing, and repeated image screening processing on the captured image set to obtain a calibration image set that ultimately contains complete, clear, non-repetitive and valid pictures, thereby ensuring that the images in the calibration image set are all suitable for the camera's intrinsic parameter calibration, which can effectively improve the calibration accuracy and efficiency.

The calibration image screening method, wherein the calibration image screening method further comprises a step performed between step S2 and step S3:

SA. grayscale the images in the captured image set.

In this example, the present application preferably performs grayscale processing on all images in the captured images before executing step S3, so that these images are converted into grayscale images to reduce the influence of other noise on corner point detection and calculation of image variance.

The calibration image screening method, wherein step S3 comprises:

S31. Acquire candidate corner points of each image in the captured image set based on a sliding window and a corner point detection algorithm.

In this example, the calibration image screening method of the present application detects corner points in the image based on a sliding window, and cooperates with the corresponding corner point detection algorithm to gradually detect the corner points in the image according to the window sliding direction.

The calibration image screening method, wherein step S5 comprises:

S51, traversing each image in the first image set and calculating the image variance corresponding to the surrounding area of each reference corner point;

S52: Remove at least one image whose variance is smaller than a preset variance threshold from the first image set to obtain the second image set.

The calibration image screening method, wherein step S6 comprises:

S61, establishing a corner point position list;

S62, traverse and determine whether the positions of the reference corner points of the edges in the images in the second image set are repeated with the positions of the reference corner points of the edges that have been placed in the corner point position list, and place the positions of the corresponding reference corner points that are not repeated into the corner point position list and remove images with repeated positions from the second image set.

In the calibration image screening method, the positional relationship of the reference corner points of the edge includes the coordinate distance of the first reference corner point and the coordinate distance of the last reference corner point.

In a second aspect, the present application further provides a calibration method, which performs camera calibration based on a calibration image set obtained by screening the calibration image screening method provided in the first aspect.

The calibration method of the present application is used to perform intrinsic parameter calibration of a camera. Based on the calibration image screening method provided in the first aspect, the captured image set is sequentially subjected to incomplete image screening processing, blurred image screening processing, and repeated image screening processing, so as to obtain a calibration image set that ultimately contains complete, clear, and non-repetitive valid images, thereby ensuring that the images in the calibration image set are all suitable for the intrinsic parameter calibration of the camera, so that the calibration method of the present application has the advantages of high calibration accuracy and efficiency.

In a third aspect, the present application further provides a calibration image screening device for screening images that can be used for camera calibration from video data, wherein the camera calibration is performed based on a checkerboard calibration plate, and the calibration image screening device comprises:

An acquisition module, used to acquire a captured video containing a checkerboard calibration plate;

An extraction module, used for extracting a set of captured images from the captured video;

A corner point detection module, used to obtain candidate corner points in each image in the captured image set;

A first screening module, configured to screen and obtain a first image set according to a distribution rule of corner points of the checkerboard calibration plate and determine reference corner points in each image in the first image set from candidate corner points, wherein the images in the first image set are images in the captured image set that include the checkerboard calibration plate;

A second screening module, configured to remove blurred images in the first image set according to the image variance of the area around the reference corner point to screen out a second image set;

The third screening module is used to remove duplicate images in the second image set according to the positional relationship of the reference corner points of the edge to screen out the calibration image set.

The calibration image screening device of the present application sequentially performs incomplete image screening processing, blurred image screening processing, and repeated image screening processing on the captured image set to obtain a calibration image set that ultimately contains complete, clear, non-repetitive and valid images, thereby ensuring that the images in the calibration image set are all suitable for the camera's intrinsic parameter calibration, which can effectively improve the calibration accuracy and efficiency.

In a fourth aspect, the present application also provides an electronic device, comprising a processor and a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the calibration image screening method provided in the first aspect or the steps in the calibration method provided in the second aspect are executed.

In a fifth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, runs the steps in the calibration image screening method provided in the first aspect or the calibration method provided in the second aspect.

From the above, it can be seen that the present application provides a calibration image screening method, a calibration method and related equipment, wherein the calibration image screening method sequentially performs incomplete image screening processing, blurred image screening processing, and repeated image screening processing on the captured image set to obtain a calibration image set that ultimately contains complete, clear, non-repetitive and valid images, ensuring that the images in the calibration image set are all suitable for the camera's intrinsic parameter calibration, which can effectively improve the calibration accuracy and efficiency.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a flow chart of a calibration image screening method provided in an embodiment of the present application.

FIG. 2 is a schematic diagram of an image containing a clear checkerboard calibration plate.

FIG. 3 is a schematic diagram of an image including a clear checkerboard calibration plate and having reference corner points marked.

FIG. 4 is a schematic diagram of an image containing a blurred checkerboard calibration plate.

FIG. 5 is a schematic diagram of the structure of a calibration image screening device provided in an embodiment of the present application.

FIG6 is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application.

Figure numerals: 301, acquisition module; 302, extraction module; 303, corner point detection module; 304, first screening module; 305, second screening module; 306, third screening module; 401, processor; 402, memory; 403, communication bus.

DETAILED DESCRIPTION

The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.

It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

In the first aspect, referring to FIG. 1 , some embodiments of the present application provide a calibration image screening method for screening images that can be used for camera calibration from video data, wherein the camera calibration is performed based on a checkerboard calibration plate, and the calibration image screening method comprises the following steps:

S1, obtaining a captured video containing a checkerboard calibration plate;

S2, extracting a set of captured images from the captured video;

S3, obtaining candidate corner points in each image in the captured image set;

S4, selecting and acquiring a first image set according to the distribution rule of the corner points of the checkerboard calibration plate and determining the reference corner points in each image in the first image set from the candidate corner points, wherein the images in the first image set are images in the captured image set that include the checkerboard calibration plate;

S5, removing blurred images in the first image set according to the image variance of the area around the reference corner point to select the second image set;

S6. Remove duplicate images in the second image set according to the positional relationship of the reference corner points of the edge to screen out a calibration image set.

Specifically, the calibration image screening method of the embodiment of the present application is applied to the camera intrinsic parameter calibration. The camera can be mounted on various mobile devices, such as manipulators, mobile robots, drones, etc., so that the corresponding mobile device can obtain the spatial position of the corresponding visual object or the mobile device itself according to the video or image captured by the camera; in the embodiment of the present application, the captured video is video data collected during the movement of the camera or during the movement of the checkerboard calibration board, and is a video segment of the checkerboard calibration board captured in the camera vision, so that the camera can perform intrinsic parameter calibration based on the position of the checkerboard calibration board with a known spatial position in the captured video, wherein the checkerboard calibration board is a board surface member with a square-shaped black and white checkerboard texture for positioning calibration.

More specifically, in order to ensure the accuracy of the camera intrinsic calibration, the calibration image screening method of the embodiment of the present application can screen out a calibration image set based on images containing a complete and clear checkerboard calibration plate from the captured video, so as to optimize the screening process of effective images for calibration.

More specifically, the extraction behavior in step S2 can be random extraction, frame-by-frame extraction, or frame-skipping extraction. Considering that the position differences of images with similar time sequences in the captured video are small, using more valid pictures with position differences for camera intrinsic parameter calibration can effectively improve the calibration accuracy. Therefore, in the embodiment of the present application, step S2 is preferably: extracting a captured image set from the captured video by skipping frames, and further preferably extracting one frame of video from every 2 or 3 frames in the captured video to form a captured image set. In addition, using frame-skipping to extract the captured image set can also improve the processing efficiency of the calibration image screening method in the embodiment of the present application.

More specifically, the background of the checkerboard calibration plate is white, and a plurality of black square grids are arranged crosswise and at intervals therein, wherein corner points are generated at the corner positions of the black square grids.

More specifically, the candidate corner points are corner points determined by preliminary analysis and detection of the corresponding image based on a corner detection algorithm.

More specifically, the calibration image screening method of the embodiment of the present application performs incomplete image screening processing, blurred image screening processing, and repeated image screening processing on the captured image set based on steps S4, S5, and S6, respectively, to obtain a calibration image set that ultimately contains complete, clear, and non-repetitive valid images.

More specifically, since the corner points in the checkerboard calibration plate are generated based on the black square grids, the distribution and number of these corner points are related to the distribution pattern of the black square grids and have obvious matrix distribution characteristics. Therefore, step S4 can determine the integrity of the checkerboard calibration plate in the corresponding image by analyzing the distribution and number of the corner points of each image in the captured image set. Therefore, step S4 can quickly determine the integrity of the checkerboard calibration plate in the corresponding image in the captured image set based on the candidate corner points detected in step S3, so as to screen out images with a complete checkerboard calibration plate to form a first image set.

More specifically, the candidate corner points are all the corner points detected in the corresponding image, and the reference corner points are the corner points selected from the candidate corner points that best highlight the distribution pattern of the corner points in the checkerboard calibration plate; in this embodiment of the present application, the reference corner points are preferably the corner points generated at the intersection of the corners of different black square grids in the checkerboard calibration plate, as shown in Figures 2 and 3, the 10 columns and 7 rows of black square grids in the figure generate 9×6=54 reference corner points.

More specifically, the reference corner point includes corner point position information, which represents the position of the reference corner point in the image, and is preferably coordinate data of the reference corner point in the image.

More specifically, since the checkerboard calibration plate is formed by a cross-spaced arrangement of black squares, there should be a clear black and white interval area near the reference corner point, and the calculation method of the image variance is mainly used to measure the degree of deviation between the pixel value in the image and its mean, thereby reflecting the grayscale richness and information content of the image. Therefore, the surrounding area of the reference corner point should have a larger image variance. If the image in the first image set is blurred, the image variance of the surrounding area of the reference corner point in the corresponding image will decrease, so that the calibration image screening method of the embodiment of the present application can quickly determine the blurriness of the corresponding image based on the image variance of the surrounding area of the reference corner point, so as to remove the blurred image from the first image set to screen out the second image set containing only clear images, wherein the image variance can be calculated using the existing regional image variance calculation formula, which will not be repeated here.

More specifically, the reference corner point of the edge corresponds to the edge position of the black square grid at the edge position of the entire checkerboard calibration plate, which can reflect the position and scaling size of the checkerboard calibration plate in the corresponding image. In the embodiment of the present application, step S6 preferably considers at least two reference corner points corresponding to the corner positions of the checkerboard calibration plate to determine whether the position and size of the checkerboard calibration plate in different images in the second image set are the same to determine whether there are checkerboard calibration plates with repeated positions and sizes in these images, so as to remove these images with repeated positions and sizes from the second image set to obtain a calibration image set. Removing these repeated images can reduce the amount of analysis for subsequent camera calibration, thereby improving calibration efficiency.

The calibration image screening method of the embodiment of the present application sequentially performs incomplete image screening processing, blurred image screening processing, and repeated image screening processing on the captured image set to obtain a calibration image set that ultimately contains complete, clear, non-repetitive and valid pictures, thereby ensuring that the images in the calibration image set are all suitable for the camera's intrinsic parameter calibration, which can effectively improve the calibration accuracy and efficiency.

It should be noted that the screening process of step S2 to step S6 may be performed by continuously extracting and analyzing images in the video during the video shooting process, or may be performed based on the video shot in advance.

In some preferred embodiments, the calibration image screening method further includes the following steps performed between step S2 and step S3:

SA, grayscale the images in the captured image set.

Specifically, although the checkerboard calibration plate is composed of a white background and black squares, other colors may be introduced into the actually acquired captured video due to factors such as scene reflection. In order to improve the accuracy of candidate corner point detection and blurred image removal, the embodiment of the present application preferably performs grayscale processing on all images in the captured image before executing step S3, so that these images are converted into grayscale images to reduce the influence of other noise on corner point detection and image variance calculation.

It should be noted that the images in the calibration image set can be images that have been grayscaled in step SA, or images that have been returned to their original colors. In the embodiment of the present application, the former is preferred, and the checkerboard calibration plate in the grayscaled image is more conducive to the calibration of the camera intrinsic parameters.

In some preferred embodiments, step S3 comprises:

S31. Acquire candidate corner points of each image in the captured image set based on a sliding window and a corner point detection algorithm.

Specifically, the corner detection algorithm is an existing algorithm for corner recognition and positioning based on an image. In the embodiment of the present application, the Harris algorithm or the Shi-Tomasi algorithm may be used.

More specifically, the calibration image screening method of the embodiment of the present application detects corner points in the image based on a sliding window, and in conjunction with a corresponding corner point detection algorithm, can gradually detect the corner points in the image according to the window sliding direction, and generate scores (response values) about the corner points based on the corresponding corner point detection algorithm, and determine the corner points with scores greater than a preset score threshold as candidate corner points, and can also determine n (set according to the black square grids in the checkerboard calibration plate) corner points with the maximum score in the image as candidate corner points.

More specifically, the candidate corner points determined in the image include the corner points of the black square grids. The image variance of the surrounding areas of the outer corner points of the outermost black square grids (not the corner points generated at the intersection of the corners of different black square grids) is small, which is not conducive to the recognition of blurred images. Therefore, step S4 needs to determine the reference corner points that are more conducive to blurred image screening from the candidate corner points according to the corner point distribution law.

In some preferred embodiments, step S5 comprises:

S51, traversing each image in the first image set and calculating the image variance corresponding to the surrounding area of each reference corner point;

S52: remove at least one image whose variance is smaller than a preset variance threshold from the first image set to obtain a second image set.

Specifically, FIG4 shows a blurred image that needs to be screened out. Based on FIG3, it can be seen that the edges of the black square grid near the reference corner point are clearly visible, and the image variance calculated in the surrounding area of the reference corner point is significantly larger than the image variance corresponding to the blurred image. Therefore, based on the image variance, it can be quickly determined whether there is a blur problem in the surrounding area of different reference corner points.

More specifically, in order to more accurately eliminate blurred images, the calibration image screening method of the embodiment of the present application performs image variance calculation on the surrounding areas of each benchmark corner point of each image in the first image set, and removes images in which at least one image variance is less than a preset variance threshold, so as to filter out all locally blurred or overall blurred images.

More specifically, the size of the peripheral area can be set according to usage requirements, and is generally set to be smaller than the size of the black square grid in vision. For example, if the image area is set to k*k, the value of k is set to be smaller than the coordinate distance between adjacent reference corner points.

In some preferred embodiments, step S6 comprises:

S61, establishing a corner point position list;

S62, traverse and determine whether the positions of the reference corner points of the edges in the images in the second image set are repeated with the positions of the reference corner points of the edges that have been placed in the corner point position list, and place the positions of the corresponding reference corner points that are not repeated into the corner point position list and remove the images with repeated positions from the second image set.

Specifically, in step S62, the process of determining whether the position of the reference corner point is repeated is a grouping determination process, that is, the reference corner points of all edges that need to be analyzed in the image are regarded as a group of data to be compared with the data in each image in the corner point position list. When a group of data is repeated, the image is regarded as a repeated image.

More specifically, in order to simplify the analysis and judgment process, the reference corner points of the edges that need to be analyzed in each image are defined as two reference corner points located at the outermost sides of the chessboard calibration plate and set diagonally, preferably the reference corner point at the outermost edge of the upper left corner (hereinafter referred to as the first reference corner point) and the reference corner point at the outermost edge of the lower right corner (hereinafter referred to as the last reference corner point). The positions of the first reference corner point and the last reference corner point can comprehensively reflect the size and position of the chessboard calibration plate in the image. Therefore, by analyzing the position data of the first reference corner point and the last reference corner point, it is possible to distinguish whether the two images are repeated.

More specifically, in this embodiment, step S62 is equivalent to determining whether the positions of the first reference corner point and the last reference corner point in the image to be analyzed are simultaneously the same as the positions of the first reference corner point and the last reference corner point corresponding to a certain image in the corner point position list, thereby determining whether the image to be analyzed is a repeated image. This determination method can quickly complete the identification of repeated images.

It should be noted that after the corner point position list is established, the corner point position list is initially empty. The first image analyzed in step S62 must be a non-repetitive image, so the position of the corresponding reference corner point can be directly placed in the corner point position list.

In some preferred embodiments, the positional relationship of the reference corner points of the edge includes the coordinate distance of the first reference corner point and the coordinate distance of the last reference corner point.

Specifically, in this embodiment, step S6 can determine whether the images are repeated by judging the size of the coordinate distance between the first reference corner point and the coordinate distance between the last reference corner points on different images in the second image set.

More specifically, step S62 can be performed by calculating the coordinate distance between the first reference corner point in the image to be analyzed and the first reference corner point of an image in the corner point position list, as well as the coordinate distance between the last reference corner point and the last reference corner point of the corresponding image in the corner point position list, and then judging whether the image to be analyzed is repeated with the images in the corner point position list based on the sizes of the two coordinate distances, and then judging whether the image to be analyzed is not repeated with the images corresponding to the positions of the reference corner points of the edges stored in the corner point position list.

More specifically, the calibration image screening method of the embodiment of the present application characterizes the positional relationship of the reference corner points of the edge based on the coordinate distance, which can further simplify the judgment process of the repeated images and improve the image screening effect.

In some more preferred embodiments, in step S62, the process of determining whether the position of the reference corner point of the edge in the image in the second image set overlaps with the position of the reference corner point of the edge placed in the corner point position list is as follows:

S621, selecting an image from the second image set as the image to be analyzed, calculating the total coordinate distance between the image to be analyzed and the image corresponding to the positions of the reference corner points of the edges placed in the corner point position list, the total coordinate distance being the sum of the coordinate distance between the image to be analyzed and the first reference corner point and the coordinate distance between the last reference corner point of the corresponding image;

S622: When there is at least one coordinate total distance less than a preset distance threshold, the image to be analyzed is considered to be a repeated image, and the position of the reference corner point of its edge is repeated with the position of the reference corner point of the edge placed in the corner point position list.

Specifically, in actual operation, there are rarely completely identical images in the second image set, but similar images are of little significance to the camera's intrinsic calibration and will reduce the calibration efficiency. Therefore, the calibration image screening method of the embodiment of the present application sets a preset distance threshold to determine whether the image is a duplicate image, regards highly similar images as duplicate images, and eliminates these duplicate images.

In some other implementations, step S6 may also include:

S61', establishing a calibration image set;

S62', traverse and determine whether the positions of the reference corner points of the edges in the images in the second image set are repeated with the positions of the reference corner points of the edges in the images already placed in the calibration image set, and place the images whose positions are not repeated into the calibration image set.

In a second aspect, some embodiments of the present application further provide a calibration method, which performs camera calibration based on a calibration image set obtained by screening the calibration image screening method provided in the first aspect.

The calibration method of the embodiment of the present application is used to perform intrinsic parameter calibration of the camera. Based on the calibration image screening method provided in the first aspect, it sequentially performs incomplete image screening processing, blurred image screening processing, and repeated image screening processing on the captured image set to obtain a calibration image set that ultimately contains complete, clear, and non-repetitive valid images, ensuring that the images in the calibration image set are all suitable for the intrinsic parameter calibration of the camera, so that the calibration method of the embodiment of the present application has the advantages of high calibration accuracy and efficiency.

In a third aspect, referring to FIG. 5 , some embodiments of the present application further provide a calibration image screening device for screening images that can be used for camera calibration from video data, wherein the camera calibration is performed based on a checkerboard calibration plate, and the calibration image screening device includes:

An acquisition module 301 is used to acquire a captured video including a checkerboard calibration plate;

An extraction module 302 is used to extract a set of captured images from a captured video;

Corner point detection module 303, used to obtain candidate corner points in each image in the captured image set;

A first screening module 304 is used to screen and obtain a first image set according to the distribution rule of corner points of the checkerboard calibration plate and determine the reference corner points in each image in the first image set from the candidate corner points, wherein the images in the first image set are images in the captured image set that include the checkerboard calibration plate;

A second screening module 305 is used to remove blurred images in the first image set according to the image variance of the area around the reference corner point to screen out the second image set;

The third screening module 306 is used to remove duplicate images in the second image set according to the positional relationship of the reference corner points of the edge to screen out the calibration image set.

The calibration image screening device of the embodiment of the present application sequentially performs incomplete image screening processing, blurred image screening processing, and repeated image screening processing on the captured image set to obtain a calibration image set that ultimately contains complete, clear, non-repetitive and valid pictures, thereby ensuring that the images in the calibration image set are all suitable for the camera's intrinsic parameter calibration, which can effectively improve the calibration accuracy and efficiency.

In some preferred embodiments, the calibration image screening device of the embodiment of the present application is used to execute the calibration image screening method provided in the first aspect above.

In the fourth aspect, please refer to Figure 6. Some embodiments of the present application also provide a structural diagram of an electronic device. The present application provides an electronic device, including: a processor 401 and a memory 402. The processor 401 and the memory 402 are interconnected and communicate with each other through a communication bus 403 and/or other forms of connection mechanisms (not marked). The memory 402 stores computer-readable instructions executable by the processor 401. When the electronic device is running, the processor 401 executes the computer-readable instructions to execute the method in any optional implementation method of the above-mentioned embodiments.

In a fifth aspect, an embodiment of the present application provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the method in any optional implementation of the above embodiment is executed. The computer-readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, disk or optical disk.

In summary, the embodiments of the present application provide a calibration image screening method, a calibration method and related equipment, wherein the calibration image screening method sequentially performs incomplete image screening processing, blurred image screening processing, and repeated image screening processing on the captured image set to obtain a calibration image set that ultimately contains complete, clear, and non-repetitive valid images, ensuring that the images in the calibration image set are all suitable for the camera's intrinsic parameter calibration, which can effectively improve the calibration accuracy and efficiency.

In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

In addition, the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

Furthermore, the functional modules in the various embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.

In this document, relational terms such as first and second, etc. are used merely to distinguish one entity or operation from another entity or operation, but do not necessarily require or imply any such actual relationship or order between these entities or operations.

The above description is only an embodiment of the present application and is not intended to limit the protection scope of the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims (7)

1. A calibration image screening method, characterized in that it is used to screen images that can be used for camera calibration from video data, wherein the camera calibration is performed based on a checkerboard calibration plate, and the calibration image screening method comprises the following steps: S1, obtaining a captured video containing a checkerboard calibration plate; S2, extracting a set of captured images from the captured video; S3, obtaining candidate corner points in each image in the captured image set; S4, selecting and acquiring a first image set according to the distribution rule of the corner points of the checkerboard calibration plate and determining the reference corner points in each image in the first image set from the candidate corner points, wherein the images in the first image set are the images in the captured image set that include the checkerboard calibration plate; S5, removing blurred images in the first image set according to the image variance of the area around the reference corner point to select a second image set; S6, according to the positional relationship of the edge reference corner points, the duplicate images in the second image set are removed to screen out the calibration image set, where the positional relationship of the edge reference corner points includes the coordinate distance of the first reference corner point and the coordinate distance of the last reference corner point; Step S5 includes: S51, traversing each image in the first image set and calculating the image variance corresponding to the surrounding area of each reference corner point; S52, removing at least one image in the first image set where the image variance is less than a preset variance threshold, to obtain the second image set; Step S6 includes: S61, establishing a corner point position list; S62, traversing and determining whether the positions of the reference corner points of the edges in the images in the second image set are repeated with the positions of the reference corner points of the edges that have been placed in the corner point position list, placing the positions of the reference corner points whose corresponding positions are not repeated into the corner point position list, and removing the images whose positions are repeated from the second image set; Step S62 includes: S621, selecting an image from the second image set as the image to be analyzed, calculating the total coordinate distance between the image to be analyzed and the image corresponding to the positions of the reference corner points of the edges placed in the corner point position list, the total coordinate distance being the sum of the coordinate distance between the image to be analyzed and the first reference corner point and the coordinate distance between the last reference corner point of the corresponding image; S622: When there is at least one coordinate total distance less than a preset distance threshold, the image to be analyzed is considered to be a repeated image, and the position of the reference corner point of its edge is repeated with the position of the reference corner point of the edge placed in the corner point position list. 2. The calibration image screening method according to claim 1, characterized in that the calibration image screening method further comprises a step performed between step S2 and step S3: SA. grayscale the images in the captured image set. 3. The calibration image screening method according to claim 1, characterized in that step S3 comprises: S31. Acquire candidate corner points of each image in the captured image set based on a sliding window and a corner point detection algorithm. 4. A calibration method, characterized in that the calibration method performs camera calibration based on a calibration image set screened and obtained by the calibration image screening method according to any one of claims 1 to 3. 5. A calibration image screening device, characterized in that it is used to screen images that can be used for camera calibration from video data, wherein the camera calibration is performed based on a checkerboard calibration plate, and the calibration image screening device comprises: An acquisition module, used to acquire a captured video containing a checkerboard calibration plate; An extraction module, used for extracting a set of captured images from the captured video; A corner point detection module, used to obtain candidate corner points in each image in the captured image set; A first screening module, configured to screen and obtain a first image set according to a distribution rule of corner points of the checkerboard calibration plate and determine reference corner points in each image in the first image set from candidate corner points, wherein the images in the first image set are images in the captured image set that include the checkerboard calibration plate; A second screening module, configured to remove blurred images in the first image set according to the image variance of the area around the reference corner point to screen out a second image set; A third screening module, for removing duplicate images in the second image set to screen out a calibration image set according to the positional relationship of the edge reference corner points, wherein the positional relationship of the edge reference corner points includes a coordinate distance of a first reference corner point and a coordinate distance of a last reference corner point; The step of removing the blurred images in the first image set according to the image variance of the area around the reference corner point to filter out the second image set comprises: S51, traversing each image in the first image set and calculating the image variance corresponding to the surrounding area of each reference corner point; S52, removing at least one image in the first image set where the image variance is less than a preset variance threshold, to obtain the second image set; The step of removing duplicate images in the second image set according to the positional relationship of the reference corner points of the edge to screen out the calibration image set comprises: S61, establishing a corner point position list; S62, traversing and determining whether the positions of the reference corner points of the edges in the images in the second image set are repeated with the positions of the reference corner points of the edges that have been placed in the corner point position list, placing the positions of the reference corner points whose corresponding positions are not repeated into the corner point position list, and removing the images whose positions are repeated from the second image set; Step S62 includes: S621, selecting an image from the second image set as the image to be analyzed, calculating the total coordinate distance between the image to be analyzed and the image corresponding to the positions of the reference corner points of the edges placed in the corner point position list, the total coordinate distance being the sum of the coordinate distance between the image to be analyzed and the first reference corner point and the coordinate distance between the last reference corner point of the corresponding image; S622: When there is at least one coordinate total distance less than a preset distance threshold, the image to be analyzed is considered to be a repeated image, and the position of the reference corner point of its edge is repeated with the position of the reference corner point of the edge placed in the corner point position list. 6. An electronic device, characterized in that it includes a processor and a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the calibration image screening method according to any one of claims 1 to 3 or the steps in the calibration method according to claim 4 are executed. 7. A computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the steps in the calibration image screening method according to any one of claims 1 to 3 or the calibration method according to claim 4 are executed.