CN111815685B - A checkerboard corner point positioning method, device and electronic device - Google Patents

Abstract

The invention provides a checkerboard angular point positioning method, a checkerboard angular point positioning device and electronic equipment, which relate to the technical field of image measurement and comprise the following steps: acquiring a target image with checkerboard characteristics; the method comprises the steps that a target image is translated by a preset number of pixels from an initial position along a target direction relative to a preset template; and taking the target images after each movement as current images one by one, and performing the following matching operation on the current images: determining a current pixel parameter based on a preset template and a current image; the current pixel parameter is used for representing the pixel information at the starting boundary and the intersection of the checkerboard in the current image; determining the matching degree between the current image and a preset template according to a correlation coefficient algorithm and the current pixel parameter; and determining the corner points of the checkerboard based on the matching degree corresponding to each current image. The method and the device can effectively reduce the data calculation amount and improve the matching efficiency and the positioning efficiency of the checkerboard angular points.

Description

A checkerboard corner point positioning method, device and electronic device

technical field

The present disclosure relates to the technical field of image measurement, and in particular, to a method, device and electronic device for locating corner points of a checkerboard.

Background technique

With the rapid development of image acquisition hardware and software, digital image correlation algorithms have gradually replaced traditional measurement methods. In particular, long-distance in-plane small displacement measurement has important applications in large-scale building structures, complex spaces, long-span bridges and other fields. . In digital image measurement, measurement speed and measurement accuracy are very important.

In the process of digital image measurement, the checkerboard with obvious texture features is mainly used to locate the corner points to realize the camera calibration. At present, the one-time matching process of the template and the real image needs to calculate a large amount of data, such as the pixel average value _Im in the sub-region of the image, in order to finally locate the corner points of the checkerboard; when the whole image is calculated, the huge amount of data and the complex complexity The calculation will lead to a drop in efficiency, which will seriously affect the speed of digital image measurement.

SUMMARY OF THE INVENTION

In order to solve the above technical problems or at least partially solve the above technical problems, the present disclosure provides a checkerboard corner point positioning method, device and electronic device.

An embodiment of the present disclosure provides a method for locating corner points of a checkerboard, including: acquiring a target image with checkerboard features; relative to a preset template, translating the target image by a preset number from an initial position along a target direction take the target images after each movement as the current image one by one, and perform the following matching operation on the current image: determine the current pixel parameter based on the preset template and the current image; wherein, the current pixel parameter is used for Indicates the pixel information at the starting boundary and the junction of the checkerboard in the current image; and, determining the degree of matching between the current image and the preset template according to the correlation coefficient algorithm and the current pixel parameter; based on The matching degree corresponding to each of the current images determines the corner points of the checkerboard.

Further, the step of determining the current pixel parameter based on the preset template and the current image includes: at a starting boundary in the current image, calculating the pixel average value of the current image based on the current image, based on The current image calculates the pixel square sum of the current image, and, based on the preset template and the current image, calculates the sum of the first pixel products between the current image and the preset template; At the junction of the checkerboard in the image, calculate the sum of the second pixel products between the current image and the preset template based on the preset template and the current image; The sum of the products of one pixel and the sum of the products of the second pixel is determined as the current pixel parameter.

Further, the expression of the correlation coefficient algorithm is:

为所述预设模板的像素方差，所述预设模板的大小尺寸为(2M+1)×(2M+1)。Wherein, C is the degree of matching between the current image and the preset template, H- _M is the sum of the first pixel products, H ₀ is the second pixel product sum, and Hr is the current image except for the In addition to the junction of the initial border and the checkerboard, the sum of the third pixel products between the current image and the preset template; I- _M is the average value of pixels at the initial border in the current image, and Ir is the The average value of pixels except the starting boundary in the current image, K- _M is the sum of squares of pixels at the starting boundary in the current image, and Kr is the sum of squares of pixels other than the starting boundary in the current image , T _m is the pixel mean value of the preset template,

is the pixel variance of the preset template, and the size of the preset template is (2M+1)×(2M+1).

Further, before performing the matching operation on the current image, the method further includes: constructing a matrix of the preset template; and calculating a pixel average value and a pixel variance of the preset template based on the matrix of the preset template.

Further, the pixel average value of the preset template is:

Wherein, T _m the pixel average value of the preset template, i represents the row number of the preset template, j represents the column number of the preset template, and T(i,j) is the number of the preset template in the The pixel value of row i and column j.

Further, the pixel variance of the preset template:

为所述预设模板的像素方差，T(i,j)为所述预设模板中第i行第j列的像素值，T_m所述预设模板的像素平均值。in,

is the pixel variance of the preset template, T(i,j) is the pixel value of the ith row and jth column in the preset template, and _Tm is the pixel average value of the preset template.

Further, the step of obtaining a target image with checkerboard features includes: obtaining a first image with checkerboard features; determining a first pixel parameter based on a preset template and the first image; according to the correlation coefficient algorithm and the first pixel parameter to determine the first matching degree between the first image and the preset template; rotate the first image by a preset angle to obtain a second image; determine the second image and the preset template The second matching degree between the preset templates; by comparing the first matching degree and the second matching degree, the image with a higher matching degree in the first image and the second image is determined as having the checkerboard feature target image.

Further, the method further includes: setting initial pixel values of different areas in the initial template; wherein the initial template includes four black and white rectangular areas, and the initial pixel values of the diagonal areas are equal; according to the target image The pixel value adjusts the initial pixel value of the initial template, and the template whose pixel value is adjusted is used as the preset template.

The embodiment of the present disclosure also provides a checkerboard corner point positioning device, including: an image acquisition module, used to obtain a target image with checkerboard characteristics; an image translation module, used to compare the target image with a preset template. Translate a preset number of pixels along the target direction from the initial position; the operation execution module is used to take the target images one by one as the current image after each movement, and perform the following matching operations on the current image: based on the preset template and For the current image, determine the current pixel parameter; wherein, the current pixel parameter is used to represent the pixel information at the starting boundary and the junction of the checkerboard in the current image; and, according to the correlation coefficient algorithm and the current pixel The parameter determines the matching degree between the current image and the preset template; the checkerboard corner point positioning module is configured to determine the checkerboard corner point based on the matching degree corresponding to each of the current images.

An embodiment of the present disclosure further provides an electronic device, the electronic device includes: a processor; a memory for storing instructions executable by the processor; the processor for reading the memory from the memory The instructions can be executed, and the instructions can be executed to realize the above-mentioned method for locating the corner points of the checkerboard.

This embodiment provides a method, device, and electronic device for locating corner points of a checkerboard. First, a target image with checkerboard features is acquired; then, the target image is moved relative to a preset template, and the target images after each movement are used as the current target image one by one. image, and perform the following matching operations on the current image: based on the preset template and the current image, determine the current pixel parameters at the starting border and the junction of the checkerboard in the current image, and determine the current image and the current image according to the correlation coefficient algorithm and the current pixel parameters. The matching degree between the preset templates; finally, based on the matching degree corresponding to each of the current images, the corner points of the checkerboard are determined. In the above checkerboard corner positioning method, with the movement of the image, each matching process between the template and the image only needs to calculate the pixel information at the boundary and the junction, which effectively reduces the amount of data calculation and improves the matching efficiency and checkerboard angle. point positioning efficiency.

Description of drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the disclosure and together with the description serve to explain the principles of the disclosure.

In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in the prior art, the accompanying drawings that are required to be used in the description of the embodiments or the prior art will be briefly introduced below. In other words, on the premise of no creative labor, other drawings can also be obtained from these drawings.

1 is a schematic flowchart of a method for locating corner points of a checkerboard provided by an embodiment of the present disclosure;

2 is a schematic diagram of a checkerboard provided by an embodiment of the present disclosure;

FIG. 3 is a schematic diagram of image movement provided by an embodiment of the present disclosure;

4 is a schematic diagram of a template provided by an embodiment of the present disclosure;

FIG. 5 is a structural block diagram of a checkerboard corner point positioning device according to an embodiment of the present disclosure.

Detailed ways

In order to more clearly understand the above objects, features and advantages of the present disclosure, the present invention will be described in detail below with reference to specific embodiments. It should be noted that the embodiments of the present disclosure and the features in the embodiments may be combined with each other under the condition of no conflict.

Many specific details are set forth in the following description to facilitate a full understanding of the present disclosure, but the present disclosure can also be implemented in other ways different from those described herein; obviously, the embodiments in the specification are only a part of the embodiments of the present disclosure, and Not all examples.

This embodiment provides a method for locating corner points of a checkerboard, and the method can be executed by a device for locating corner points of a checkerboard, wherein the device can be implemented by software and/or hardware, and can generally be integrated into an electronic device. Referring to the flowchart of the method for locating the corner points of the checkerboard shown in FIG. 1, the method includes:

Step S102, acquiring a target image with checkerboard features.

In this embodiment, an object with checkerboard characteristics (such as a calibration board) can be used as a test object, and the test object is photographed by a camera to obtain a target image. The target image with the checkerboard feature can be referred to as shown in FIG. 2 , and the checkerboard is composed of a plurality of black and white rectangular areas.

Step S104, with respect to the preset template, the target image is shifted by a preset number of pixels along the target direction from the initial position. The preset template is also a template with checkerboard characteristics, and reference may be made to the checkerboard shown in FIG. 2 .

In practical applications, it is possible to keep the position of the preset template unchanged and move the target image, and also to keep the position of the target image unchanged and move the preset template, thereby realizing the relative position of the target image to the preset template. move. Taking FIG. 2 as an example, the target direction may be the horizontal or vertical direction of the checkerboard, and the preset number is at least one, for example, the target image is translated by one pixel along the horizontal direction.

In step S106, the target images after each movement are taken as the current image one by one, and the matching operations shown in the following steps S1062 and S1064 are performed on the current image:

Step S1062, based on the preset template and the current image, determine the current pixel parameter; wherein, the current pixel parameter is used to represent the pixel information at the starting boundary and the junction of the checkerboard in the current image; and, step S1064, according to the correlation coefficient algorithm and the current pixel parameter to determine the matching degree between the current image and the preset template.

It can be seen from Figure 2 that the checkerboard has repeating units, each repeating unit can be called a checkerboard unit, and the checkerboard unit can be divided into four regions: I, II, III and IV. Considering that the checkerboard cell has obvious characteristics: the pixel values in the diagonal area are equal; based on this, in the process of moving the target image, for the pixels located in a certain area (such as the I area) before the target image is moved, if the target image is in the target image. After the image is moved, if the pixel is still in the same area, the matching degree between it and the preset template remains unchanged. Only the pixels in the transformed area need to recalculate the corresponding matching degree. Therefore, according to the above characteristics, this embodiment only needs to calculate the matching degree for the pixels in the transformation area. It can be understood that the transformation area is the boundary between the border and the checkerboard. Therefore, this embodiment only needs to calculate the matching degree for the starting boundary in the current image. The pixel information at the junction with the checkerboard determines the matching degree between the current image and the preset template, without the need to calculate the matching degree for other areas, which can significantly reduce the amount of matching degree calculation and effectively improve the calculation efficiency.

Step S108: Determine the corner points of the checkerboard based on the matching degree corresponding to each current image.

In this embodiment, every time the target image is moved, that is, for each current image, the pixel points whose matching degree is higher than the preset threshold (for example, 80%) can be determined as the checkerboard corner points of the current image; when When the target image is moved from the start boundary of one side to the end boundary of the opposite side, the checkerboard corner points corresponding to each current image are obtained, thereby determining all checkerboard corner points of the target image.

The method for locating the corner points of a checkerboard provided by this embodiment firstly acquires a target image with checkerboard characteristics; then moves the target image relative to a preset template, takes the target images after each movement as the current image one by one, and executes the execution on the current image. The following matching operations: based on the preset template and the current image, determine the current pixel parameters at the starting boundary and the junction of the checkerboard in the current image, and determine the matching between the current image and the preset template according to the correlation coefficient algorithm and the current pixel parameters degree; finally, based on the matching degree corresponding to each of the current images, the corner points of the checkerboard are determined. In the above checkerboard corner positioning method, with the movement of the image, each matching process between the template and the image only needs to calculate the pixel information at the boundary and the junction, which effectively reduces the amount of data calculation and improves the matching efficiency and checkerboard angle. point positioning efficiency.

In order to better understand the method for locating the corner points of a checkerboard provided in this embodiment, the method is described in detail below.

In this embodiment, it is assumed that the matrix size of the preset template is (2M+1)×(2M+1). During the matching process, the selected matching area in the target image is the same as the size of the preset template, which is also (2M+1)×(2M+1), then the correlation coefficient algorithm used to calculate the matching degree between the preset template and the target image can be shown in the following formula (1):

（1）

(1)

Among them, T(i,j) represents the pixel value of the preset template, I represents the pixel value of the target image, i represents the number of rows of the preset template, j represents the number of columns of the preset template, Δi and Δj represent the relative value of the preset template. In the relative position of the center point of the target image, T _m represents the pixel average value of the preset template, and its calculation formula is shown in (2), and I _m the pixel average value of the target image, and its calculation formula is shown in (3):

The inventor found that in the process of calculating the matching degree between the preset template and the target image according to the correlation coefficient algorithm shown in the above formula (1), each matching process needs to calculate the pixel average value _Im in the sub-region of the target image. , the pixel product T(i,j)I(i+Δi,j+Δj) of the preset template and the target image, and the pixel average value _Tm in the preset template area, the pixel I(i+ Δi,j+Δj) and sum of squares. Obviously, when calculating the entire target image, the complex calculation process of the matching degree will lead to a decrease in efficiency.

However, considering that the checkerboard cells mentioned in the foregoing embodiments have the feature that the pixel values of the diagonal regions are equal, the matching needs to be recalculated only when the target image is moved before and after the pixel points change the region. Based on this, this embodiment can provide a fast and efficient method for locating corner points of a checkerboard according to the above characteristics, as shown below.

For the preset template whose matrix size is (2 M +1)×(2 M +1), the matrix is represented as T, and the following formula (4) is used:

（4）

(4)

Among them, T _I-III and T _II-IV and respectively represent the diagonal matrix of the checkerboard, and the values in the matrix are the same.

为：Based on this, calculate the pixel variance of the preset template

for:

（5）

(5)

Wherein, T(i,j) is the pixel value of the i-th row and the j-th column in the preset template, and _Tm is the pixel average value of the preset template.

Calculate the sum H of the pixel product T(i,j)I(i+Δi,j+Δj) of the preset template and the template image as:

（6）

(6)

Among them, H - _M represents the sum of the first pixel products between the target image and the preset template at the specified boundary in the target image, and the specified boundary is the starting boundary of the target image, and -M represents the starting boundary of the target image. The column number of the border; H ₀ represents the sum of the second pixel products between the target image and the preset template at the junction of the checkerboard in the target image; Hr represents the sum of the second pixel products in the target image except the starting border and the junction of the checkerboard The area outside the target image, that is, the area in the target image with column numbers other than -M and 0, is the sum of the third pixel products between the target image and the preset template.

The pixel average value _Im of the target image is expressed as the following formula (7):

Among them, I - _M represents the pixel average value of this area at the starting boundary in the target image; Ir represents the pixel average value of other areas in the target image except the starting boundary.

The square sum K of the pixels of the target image I(i+∆i,j+∆j) is expressed as the following formula (8):

（8）

(8)

Among them, K - _M represents the pixel square sum of this area at the starting boundary in the target image; Kr represents the pixel square sum of other areas in the target image except the starting boundary.

According to the above formulas (2) to (8), the correlation coefficient algorithm shown in formula (1) can be modified as shown in the following formula (9):

（9）

(9)

为预设模板的像素方差，预设模板的大小尺寸为(2M+1)×(2M+1)。Among them, C represents the matching degree between the target image and the preset template, H - _M is the sum of the first pixel product, H ₀ is the second pixel product sum, Hr is the current image except the starting boundary and the checkerboard boundary In addition, the sum of the third pixel product between the target image and the preset template; I - _M is the pixel average value at the starting boundary in the current image, and Ir is the pixel average value other than the starting boundary in the current image. , K - _M is the square sum of pixels at the starting border in the current image, Kr is the square sum of pixels in the current image except the starting border, T _m is the pixel average value of the preset template,

is the pixel variance of the preset template, and the size of the preset template is (2 M +1)×(2 M +1).

When the algorithm shown in formula (2) is used, three columns of data need to be calculated each time, and each column contains (2M+1) data, namely H _i , I _i and K _i , a total of (6M+3) data However, using the correlation coefficient algorithm shown in formula (9), since the area where the pixel is located during the translation of the target image remains unchanged, the calculation result remains unchanged, and only new data is needed after the translation of the target image. Replacing the four columns of H _-M , H ₀ , K _-M and I _-M in the boundary area and the checkerboard boundary area, the above calculation process can be completed without updating all the data each time. Moreover, regardless of the size of the preset template, the updated data will always be the data in these four columns. Compared with the traditional calculation method, the calculation amount will increase exponentially with the increase of the template. This embodiment uses the formula shown in formula (9). The algorithm can significantly reduce the amount of calculation.

In practical applications, in order to facilitate the updating of the above-mentioned four columns of data, the present embodiment may store the above-mentioned four columns of data according to the data format; taking the scene of horizontally moving the target image as an example, the above-mentioned data format may be to store the data I , K and H , the columns of the array correspond one-to-one with the columns of the preset template. Referring to formula (10), the data is marked as A; in this case, the boundaries of the data correspond to three positions -M, 0 and M respectively. Use P to identify its starting position.

（10）

(10)

It should be noted that, since this embodiment is described by taking the target image moving in the horizontal direction as an example, in a scene where the target image moves in the vertical direction, the method for locating the corner points of the checkerboard provided in this embodiment is still applicable. , for this formula (10), the data format is to store the data I , K and H in sequence according to the columns, and the rows of the array correspond one-to-one with the rows of the preset template.

This embodiment continues to take the scene of moving the target image horizontally as an example, and in this example, the target image is moved horizontally to the left by one pixel at a time. During specific movement, the preset template can be controlled to move one pixel to the right, which is equivalent to moving the target image to the left by one pixel.

Referring to FIG. 3 , before moving the target image, the starting position of the data is marked with a mark p, and p=p ₁ is marked for the first time. After moving the target image, the column data where p ₁ is located needs to be updated, and the updated data is the data of the M+1 column of the right image. At the same time, the data H of the M+1th column starting from the starting point p ₁ needs to update its state, and then the starting point p is updated from p ₁ to p ₂ , where p ₂ =p ₁ mod(2M+1)+1, mod is to find remainder function.

In the process of moving the target image, the target images after each movement are taken as the current image one by one, and a matching operation is performed on the current image, which specifically includes:

；The step of updating the current pixel parameters: at the specified boundary in the current image, the specified boundary is a data area where the number of columns of this embodiment is -M, with reference to the above formula (7), the pixel average value I of the current image is calculated based on the current image. _-M , with reference to the above formula (8), calculate the pixel square sum K _-M of the current image based on the current image, and, with reference to the above formula (6), based on the preset template and the current image Calculate the difference between the current image and the preset template based on the preset template and the current image Sum of first pixel products

;

；At the junction of the checkerboard in the current image, the checkerboard boundary is the data area with the column number of 0 in this embodiment. Referring to the above formula (6), calculate the distance between the current image and the preset template based on the preset template and the current image. The sum of the second pixel products of

;

So far, the pixel average value I _-M , the pixel square sum K _-M , the first pixel product sum H _-M , and the second pixel product sum H ₀ calculated above are determined as the current pixel parameters.

The step of determining the matching degree between the current image and the preset template: determine the matching degree between the current image and the preset template according to the correlation coefficient algorithm shown in formula (9) and the above-mentioned current pixel parameters.

After completing the matching operations shown in the above two steps for the current image, it is necessary to update the following parameters corresponding to the current image:

表示更新后放在p位置的当前图像的像素平均值；

表示更新后放在p+1位置的当前图像的像素平均值；Ir表示更新前除了p位置和p+1位置之外，当前图像的像素平均值；I'r表示更新后除了p位置和p+1位置之外，当前图像的像素平均值；

表示更新后放在p位置的当前图像的像素平方和；K'r表示更新后除了p位置和p+1位置之外，当前图像的像素平方和；

表示更新后放在p位置的当前图像与预设模板之间的第一像素乘积之和。in,

Indicates the pixel average value of the current image placed at the p position after the update;

Represents the pixel average value of the current image placed at the p+1 position after the update; Ir represents the pixel average value of the current image before the update except for the p position and p+1 position; I'r represents the update except for the p position and p position. Outside the +1 position, the pixel average of the current image;

Indicates the square sum of the pixels of the current image placed at the p position after the update; K'r indicates the square sum of the pixels of the current image except for the p position and p+1 position after the update;

Indicates the sum of the first pixel products between the current image and the preset template after updating.

Each time the target image is moved, the matching operation described in the above embodiment is performed until all calculations of the target image are completed, the matching degree in each matching process is obtained, and the checkerboard corner position of the target image is determined accordingly.

In addition, it can be understood that the checkerboard feature has an anti-scaling function, that is to say, the change of the image size will not affect the positioning of the corner points, but the checkerboard feature cannot resist rotation. When the image is rotated, there will be a matching failure, which will affect the calculation of matching degree of efficiency and accuracy. Based on this, for the above step S102, this embodiment provides a specific method for acquiring a target image with a suitable angle, including the following steps 1 to 6:

Step 1, acquiring a first image. The first image is an actually captured image, and there may be a large angular deviation relative to the preset template.

Step 2: Determine the first pixel parameter based on the preset template and the first image. The first pixel parameter includes: at the initial boundary in the first image, the pixel average value of the first image, the pixel square sum of the first image, the first pixel product sum between the first image and the preset template, and the At the junction of the checkerboards in the first image, the sum of the second pixel products between the first image and the preset template. And the determination method of the above-mentioned first pixel parameter may refer to the aforementioned formulas (6)-(8), which will not be described further here.

Step 3: Determine the first degree of matching between the first image and the preset template according to the correlation coefficient algorithm shown in formula (9) and the first pixel parameter.

Step 4: Rotate the first image by a preset angle to obtain a second image. Specifically, based on the feature that the preset template has equal pixels in the diagonal regions, the first image can usually be rotated by a preset angle of 45°. Of course, other angles can also be rotated, and the first image can be rotated multiple times with different angles each time, so as to obtain the second image under various rotation angles.

Step 5: Determine the second degree of matching between the second image and the preset template. The specific method for determining the second matching is the same as the method for determining the first matching degree, mainly calculating the pixel parameters corresponding to the second image first, and then calculating the second matching degree in combination with the correlation coefficient algorithm, and the specific process will not be repeated.

Step 6, by comparing the first matching degree and the second matching degree, determine an image with a higher matching degree among the first image and the second image as a target image with a checkerboard feature.

Referring to the template shown in Figure 4, when the image is completely consistent with the preset template (which can be called a positive template), the corresponding matching degree is 1, and when the image is completely opposite to the template (called an anti-template), its matching degree On the contrary, the corresponding matching degree is 0. Therefore, the relative angle between the two can be determined according to the matching degree between the image and the preset template. When the matching degree is high, it means that the relative angle deviation between the two is small, so the image with the high matching degree is used as the target image. , for example, when the second matching degree is relatively high, the second image rotated by the preset angle may be determined as the target image.

In the acquisition method of the target image provided by this embodiment, by rotating the image and selecting an image with a suitable angle as the target image, the ability of the target image to resist the rotation corner point positioning can be increased, and further in the process of using the target image for corner point positioning , the correlation coefficient algorithm does not need to be adjusted, and the scope of application of the algorithm to images is increased.

In addition, for the preset template in the above embodiment, this embodiment can also provide a method for obtaining the preset template, including the following two steps:

First, set the initial pixel values of different areas in the initial template; wherein, the initial template includes four black and white rectangular areas I, II, III, IV, and the initial pixel values of the diagonal areas are equal, for example, you can set the area I and The initial pixel value of area III is 165, and the initial pixel value of area II and area IV is 30.

Then, the initial pixel value of the initial template is adjusted according to the pixel value of the target image, and the template whose pixel value is adjusted is used as the preset template. In order to reduce the calculation difficulty between the target image and the template, the initial pixel value can be adjusted according to the pixel value of the target image. Within the range of , first determine the pixel average value of the target image in each area of the initial template, and then directly adjust the initial pixel value of the area to the pixel average value of the target image corresponding to the area.

To sum up, with the method for locating the corner points of the checkerboard provided by the above embodiments, as the image moves, each matching process between the template and the image only needs to calculate the pixel information of the boundary and the junction, which effectively reduces the amount of data calculation and improves the matching process. Efficiency and localization efficiency of checkerboard corners.

This embodiment also provides a checkerboard corner point positioning device, since the checkerboard corner point positioning method provided by the above embodiment is implemented. 5, the device includes the following modules:

an image acquisition module 502, configured to acquire a target image with checkerboard features;

The image translation module 504 is used to translate the target image by a preset number of pixels along the target direction from the initial position relative to the preset template;

The operation execution module 506 is used to take the target images after each movement as the current image one by one, and perform the following matching operations on the current image:

Based on the preset template and the current image, the current pixel parameter is determined; wherein, the current pixel parameter is used to represent the pixel information at the starting boundary and the junction of the checkerboard in the current image; and, the current pixel parameter is determined according to the correlation coefficient algorithm and the current pixel parameter. The match between the image and the preset template;

The checkerboard corner point location module 508 is configured to determine the checkerboard corner points based on the matching degree corresponding to each current image.

With the checkerboard corner point positioning device provided in this embodiment, with the movement of the image, each matching process between the template and the image only needs to calculate the pixel information of the boundary and the junction, which effectively reduces the amount of data calculation and improves the matching efficiency and the checkerboard. The positioning efficiency of grid corners.

In one embodiment, the operation execution module 506 is further configured to: at the starting boundary in the current image, calculate the pixel average value of the current image based on the current image, calculate the pixel square sum of the current image based on the current image, and, based on the pre- Set the template and the current image to calculate the sum of the first pixel products between the current image and the preset template; at the junction of the checkerboard in the current image, calculate the second pixel between the current image and the preset template based on the preset template and the current image. The sum of pixel products; the calculated average value of pixels, the sum of squares of pixels, the sum of the first pixel products, and the sum of the second pixel products are determined as the current pixel parameters.

In an embodiment, the above-mentioned checkerboard corner point positioning device further includes a template matrix calculation module (not shown in the figure), which is used for: constructing a matrix of preset templates; calculating the matrix of preset templates based on the matrix of preset templates Pixel mean and pixel variance.

In an embodiment, the above image acquisition module 502 is further configured to: acquire a first image; determine a first pixel parameter based on a preset template and the first image; determine the correlation coefficient between the first image and the first pixel parameter the first matching degree between the preset templates; rotating the first image by a preset angle to obtain a second image; determining the second matching degree between the second image and the preset template; by comparing the first matching degree and the second matching degree Matching degree, the image with higher matching degree among the first image and the second image is determined as the target image with the checkerboard feature.

In an embodiment, the above-mentioned checkerboard corner point positioning device further includes a template preset module (not shown in the figure), which is used for: setting initial pixel values of different regions in the initial template; wherein, the initial template includes black and white The initial pixel value of the diagonal area is equal; the initial pixel value of the initial template is adjusted according to the pixel value of the target image, and the template after adjusting the pixel value is used as the preset template.

The implementation principle and technical effects of the device provided in this embodiment are the same as those in the foregoing method embodiments. For brief description, for the parts not mentioned in this embodiment, reference may be made to the corresponding content in the foregoing Embodiment 1.

Based on the foregoing embodiments, this embodiment provides an electronic device, including: a processor, a memory for storing executable instructions of the processor; a processor, for reading the executable instructions from the memory instruction, and execute the instruction to realize the above-mentioned method for locating the corner points of the checkerboard.

Those skilled in the art can clearly understand that, for the convenience and brevity of description, for the specific working process of the electronic device described above, reference may be made to the corresponding process in the foregoing method embodiments, which will not be repeated here.

It should be noted that, in this document, relational terms such as "first" and "second" etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply these There is no such actual relationship or sequence between entities or operations. Moreover, the terms "comprising", "comprising" or any other variation thereof are intended to encompass a non-exclusive inclusion such that a process, method, article or device that includes a list of elements includes not only those elements, but also includes not explicitly listed or other elements inherent to such a process, method, article or apparatus. Without further limitation, an element qualified by the phrase "comprising a..." does not preclude the presence of additional identical elements in a process, method, article or apparatus that includes the element.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, but not to limit the protection scope of the present invention. Persons should understand that the technical solutions of the present invention may be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention.

Claims (9)

1. a checkerboard corner point positioning method, is characterized in that, comprises: Obtain a target image with checkerboard features; With respect to the preset template, the target image is shifted from the initial position by a preset number of pixels along the target direction; Take the target images after each movement as the current image one by one, and perform the following matching operations on the current image: Based on the preset template and the current image, the current pixel parameter is determined; wherein, the current pixel parameter is used to represent the pixel information at the starting boundary and the junction of the checkerboard in the current image; and, according to the correlation coefficient algorithm and the current pixel parameter to determine the degree of matching between the current image and the preset template; Determine the corner points of the checkerboard based on the matching degree corresponding to each of the current images; The expression of the correlation coefficient algorithm is:

Wherein, C is the matching degree between the current image and the preset template, and H- _M is the first boundary between the current image and the preset template at the starting boundary in the current image. The sum of pixel products, H ₀ is the sum of the second pixel products between the current image and the preset template at the junction of the checkerboard in the current image, Hr is the current image except for the starting boundary and the area beyond the border of the checkerboard, the sum of the third pixel product between the current image and the preset template; I- _M is the pixel average value at the starting boundary in the current image, and Ir is the The average value of pixels except the starting boundary in the current image, K- _M is the sum of squares of pixels at the starting boundary in the current image, and Kr is the sum of squares of pixels other than the starting boundary in the current image , T _m is the pixel mean value of the preset template,

is the pixel variance of the preset template, the size of the preset template is ( 2M +1)×(2M+1), and M represents the sequence number of rows and/or columns of the preset template. 2. The method according to claim 1, wherein the step of determining the current pixel parameter based on the preset template and the current image comprises: at a starting boundary in the current image, calculating a pixel average value of the current image based on the current image, calculating a pixel square sum of the current image based on the current image, and, based on a preset template and the The current image calculates the sum of the first pixel products between the current image and the preset template; at the junction of checkerboards in the current image, calculating the sum of second pixel products between the current image and the preset template based on the preset template and the current image; The calculated average value of pixels, the sum of squares of pixels, the sum of first pixel products, and the sum of second pixel products are determined as the current pixel parameters. 3. The method according to claim 2, wherein before performing the matching operation on the current image, the method further comprises: constructing a matrix of said preset templates; Calculate the pixel mean and pixel variance of the preset template based on the matrix of the preset template. 4. The method according to claim 3, wherein the pixel average value of the preset template is:

Wherein, T _m the pixel average value of the preset template, i represents the row number of the preset template, j represents the column number of the preset template, and T(i,j) is the number of the preset template in the The pixel value of row i and column j. 5. The method according to claim 2, wherein the pixel variance of the preset template:

in,

is the pixel variance of the preset template, T(i,j) is the pixel value of the ith row and jth column in the preset template, and _Tm is the pixel average value of the preset template. 6. The method according to claim 1, wherein the step of acquiring a target image with checkerboard features comprises: get the first image; determining a first pixel parameter based on the preset template and the first image; determining a first degree of matching between the first image and the preset template according to the correlation coefficient algorithm and the first pixel parameter; Rotating the first image by a preset angle to obtain a second image; determining a second degree of matching between the second image and the preset template; By comparing the first matching degree and the second matching degree, an image with a higher matching degree among the first image and the second image is determined as a target image with a checkerboard feature. 7. The method according to claim 1, wherein the method further comprises: Setting the initial pixel values of different areas in the initial template; wherein, the initial template includes four black and white rectangular areas, and the initial pixel values of the diagonal areas are equal; The initial pixel value of the initial template is adjusted according to the pixel value of the target image, and the template whose pixel value is adjusted is used as the preset template. 8. A checkerboard corner point positioning device, characterized in that, comprising: an image acquisition module, used to acquire a target image with checkerboard features; an image translation module, used to translate the target image by a preset number of pixels along the target direction from the initial position relative to the preset template; The operation execution module is used to take the target images after each movement as the current image one by one, and perform the following matching operations on the current image: Based on the preset template and the current image, the current pixel parameter is determined; wherein, the current pixel parameter is used to represent the pixel information at the starting boundary and the junction of the checkerboard in the current image; and, according to the correlation coefficient algorithm and the current pixel parameter to determine the degree of matching between the current image and the preset template; a checkerboard corner point positioning module, configured to determine the checkerboard corner points based on the degree of matching corresponding to each of the current images; The expression of the correlation coefficient algorithm is:

Wherein, C is the matching degree between the current image and the preset template, and H- _M is the first boundary between the current image and the preset template at the starting boundary in the current image. The sum of pixel products, H ₀ is the sum of the second pixel products between the current image and the preset template at the junction of the checkerboard in the current image, Hr is the current image except for the starting boundary and the area beyond the border of the checkerboard, the sum of the third pixel product between the current image and the preset template; I- _M is the pixel average value at the starting boundary in the current image, and Ir is the The average value of pixels except the starting boundary in the current image, K- _M is the sum of squares of pixels at the starting boundary in the current image, and Kr is the sum of squares of pixels other than the starting boundary in the current image , T _m is the pixel mean value of the preset template,

is the pixel variance of the preset template, the size of the preset template is ( 2M +1)×(2M+1), and M represents the sequence number of rows and/or columns of the preset template. 9. An electronic device, characterized in that the electronic device comprises: processor; memory for storing instructions executable by the processor; The processor is configured to read the executable instructions from the memory, and execute the instructions to implement the method for locating a checkerboard corner point according to any one of claims 1-7 above.

Images