Billiard target rapid identification method and system

Abstract

The invention provides a billiard target rapid identification method and system. The method comprises the following steps of: acquiring a multi-angle image of a target billiard ball through a double-camera group, quickly cutting the image and performing first color gamut conversion through a machine vision platform, and performing ellipticity distortion correction to obtain a billiard ball imaging ellipticity losing true circle correction image; adding a mask in the billiard ball imaging ellipsoid missed circle correction image, carrying out first binarization processing, filtering noisy points, and obtaining a noise-filtered overall contour area of the billiard ball; second color gamut conversion and second binarization processing are carried out in the noise-filtered overall contour area of the billiard ball, a white contour area of the billiard ball is recognized, noise points in the white area are filtered, and the size of the billiard ball is calculated and recognized; and according to the billiard ball imaging ellipticization true circle correction image and the billiard ball contour white area parameters, removing the white area and filtering noisy points, carrying out equalization operation on the billiard ball contour remaining color area, and rapidly calculating and identifying the specific ball number of the target billiard ball.

Description

Technical Field

[0001] The present invention relates to the technical field of target detection, imaging, deviation correction and intelligent recognition, and more specifically, to a method and system for rapid recognition of billiard targets. Background Art

[0002] At present, with the rapid development of billiards and its intelligence, the intelligence of billiards equipment systems and intelligent referee identification methods are becoming more and more important. In the identification of multi-target ball sports targets, billiards has the characteristics of complex sports rules, small targets, and large numbers. Rapid identification of billiard targets and ball numbers is very critical; specifically, including: how to obtain multi-angle images of target billiards and detect target imaging distortion and improve cropping conversion speed, how to correct target imaging distortion, how to add a mask to remove noise and obtain the overall contour area of ​​the filtered billiards, how to identify the outline of the billiards and the size of the ball number, how to quickly calculate and identify the specific ball number of the target billiards in the billiard contour area, and other issues remain to be solved; therefore, it is necessary to propose a billiard target rapid identification method and system to at least partially solve the problems existing in the prior art. Summary of the invention

[0003] A series of simplified concepts are introduced in the content of the invention, which will be further described in detail in the specific implementation method section; the content of the invention of the present invention does not mean to attempt to limit the key features and essential technical features of the technical solution claimed for protection, nor does it mean to attempt to determine the scope of protection of the technical solution claimed for protection.

[0004] To at least partially solve the above problems, the present invention provides a method for quickly identifying a billiard target, comprising:

[0005] S100 uses a dual camera group to obtain multi-angle images of the target billiard ball, quickly crops the image and performs the first color gamut conversion through the machine vision platform, and performs elliptical distortion correction to obtain a circular correction image of the billiard ball imaging elliptical distortion;

[0006] S200, adding a mask to the circularly corrected image of the billiard ball imaging elliptical distortion, performing a first binarization process and filtering noise points, and obtaining an overall contour area of ​​the noise-filtered billiard ball;

[0007] S300, performing a second color gamut conversion and a second binarization process in the overall contour area of ​​the filtered billiard ball, identifying the white area of ​​the billiard ball contour, filtering the noise points in the white area, and calculating and identifying the size of the ball number;

[0008] S400, based on the circular deflection correction image of the billiard ball imaging elliptical distortion and the white area parameters of the billiard ball outline, removes the white area and filters the noise, performs an average operation on the remaining color area of ​​the billiard ball outline, and quickly calculates and identifies the specific ball number of the target billiard ball.

[0009] Preferably, S100 includes:

[0010] S101, using a dual camera group to capture multi-angle images of target billiard balls, and quickly cropping the multi-angle images of the target billiard balls through a machine vision platform to obtain a cropped matrix area image;

[0011] S102, performing a first color gamut conversion on the cropped matrix region image to obtain a matrix region color gamut conversion image;

[0012] S103, according to the matrix area color gamut conversion image, correcting the elliptical distortion of the billiard image in the matrix area color gamut conversion image, and obtaining a circular correction image of the elliptical distortion of the billiard image.

[0013] Preferably, S200 includes:

[0014] S201, adding a mask to the billiard ball imaging elliptical distortion circular correction image to cover the non-billiard ball area, and obtaining a billiard ball color gamut conversion mask image;

[0015] S202, by performing a first binarization process on the billiard ball color gamut conversion mask image, the overall outline of the billiard ball is identified, and noise points are filtered to obtain a noise-filtered billiard ball overall outline region.

[0016] Preferably, S300 includes:

[0017] S301, performing a second color gamut conversion in the noise-filtered billiard ball overall contour region to obtain a color gamut converted billiard ball contour region;

[0018] S302, converting the billiard ball outline area according to the color gamut, performing a second binarization process, identifying the white area of ​​the billiard ball outline and filtering the noise in the white area, and calculating and identifying the size of the ball number;

[0019] The parameters of the white area of ​​the billiard ball outline include: the shape of the white area of ​​the billiard ball outline, the perimeter of the white area of ​​the billiard ball outline and the area of ​​the white area of ​​the billiard ball outline.

[0020] Preferably, S400 includes:

[0021] S401, according to the circular deflection correction image of the billiard ball imaging elliptical distortion and the white area parameters of the billiard ball outline, the white area is removed and the noise is filtered, and only the billiard ball outline area is retained to obtain the remaining color area of ​​the billiard ball outline;

[0022] S402, performing a mean value operation on the remaining color area of ​​the billiard ball outline to obtain a remaining color average value; and quickly calculating and identifying a specific ball number of the target billiard ball based on the remaining color average value.

[0023] The present invention provides a billiard target rapid identification system, comprising:

[0024] The module for fast cutting and correction of elliptical distortion acquires multi-angle images of the target billiard ball through a dual-camera group, quickly cuts the image and performs the first color gamut conversion through a machine vision platform, and performs elliptical distortion correction to obtain a circular correction image of the elliptical distortion of the billiard ball imaging;

[0025] The mask processing noise filtering module adds a mask to the billiard ball imaging elliptical distortion circular correction image, performs the first binarization processing and filters the noise to obtain the overall contour area of ​​the noise-filtered billiard ball;

[0026] The color gamut conversion contour size recognition module performs a second color gamut conversion and a second binarization process in the overall contour area of ​​the filtered noise billiard ball, identifies the white area of ​​the billiard ball contour, filters the noise in the white area, and calculates and recognizes the size of the ball number;

[0027] The regional mean ball number recognition module removes the white area and filters the noise based on the circular correction image of the billiard ball's elliptical distortion and the white area parameters of the billiard ball's outline, performs a mean operation on the remaining color area of ​​the billiard ball's outline, and quickly calculates and identifies the specific ball number of the target billiard ball.

[0028] Preferably, the fast cutting ellipse distortion correction module includes:

[0029] The machine vision fast cutting unit uses a dual camera group to capture multi-angle images of target billiard balls, and quickly cuts the multi-angle images of target billiard balls through a machine vision platform to obtain the cut matrix area image;

[0030] A matrix area color gamut conversion unit performs a first color gamut conversion on the cropped matrix area image to obtain a matrix area color gamut conversion image;

[0031] The billiard imaging elliptical distortion correction unit performs elliptical distortion correction on the billiard image elliptical distortion matrix area in the matrix area gamut conversion image according to the matrix area gamut conversion image, and obtains the billiard imaging elliptical distortion circular correction image.

[0032] Preferably, the mask processing noise filtering module includes:

[0033] The mask adding area covering unit adds a mask to the billiard imaging elliptical distortion circular correction image to cover the non-billiard area and obtain the billiard color gamut conversion mask image;

[0034] The contour recognition noise filtering unit performs the first binarization processing on the billiard ball color gamut conversion mask image, recognizes the overall contour of the billiard ball, filters the noise, and obtains the overall contour area of ​​the noise-filtered billiard ball.

[0035] Preferably, the color gamut conversion contour size recognition module includes:

[0036] The contour area color gamut conversion unit performs a second color gamut conversion in the overall contour area of ​​the noise-filtered billiard ball to obtain a color gamut converted billiard ball contour area;

[0037] The conversion area white recognition unit converts the billiard ball outline area according to the color gamut, performs a second binarization process, identifies the white area of ​​the billiard ball outline and filters the noise in the white area, and calculates and recognizes the size of the ball number;

[0038] The parameters of the white area of ​​the billiard ball outline include: the shape of the white area of ​​the billiard ball outline, the perimeter of the white area of ​​the billiard ball outline and the area of ​​the white area of ​​the billiard ball outline.

[0039] Preferably, the regional mean ball number recognition module includes:

[0040] The residual color area filtering unit removes the white area and filters the noise according to the circular correction image of the billiard ball imaging elliptical distortion and the white area parameters of the billiard ball outline, and only retains the billiard ball outline area to obtain the residual color area of ​​the billiard ball outline;

[0041] The mean operation ball number fast identification unit performs a mean operation on the remaining color area of ​​the billiard ball outline to obtain the remaining color average value; based on the remaining color average value, the specific ball number of the target billiard ball is quickly calculated and identified.

[0042] The beneficial effects of the above technical solution include:

[0043] The present invention provides a method and system for quickly identifying a billiard target. The method obtains multi-angle images of a target billiard through a dual camera group, quickly cuts images and performs a first color gamut conversion through a machine vision platform, and performs elliptical distortion correction to obtain a circular correction image of the billiard imaging elliptical distortion; adds a mask to the circular correction image of the billiard imaging elliptical distortion, performs a first binarization process and filters noise points to obtain the overall contour area of ​​the noise-filtered billiard; performs a second color gamut conversion and a second binarization process in the overall contour area of ​​the noise-filtered billiard, identifies the white area of ​​the billiard outline, filters the noise points in the white area, and calculates and identifies the size of the ball; and calculates and identifies the size of the ball according to the circular correction image of the billiard imaging elliptical distortion and the white area of ​​the billiard outline. Parameters, remove white areas and filter noise, average the remaining color areas of the billiard ball outline, and quickly calculate and identify the specific ball number of the target billiard ball; it can obtain multi-angle images of the target billiard ball and detect the target imaging distortion and cropping conversion speed; it can correct the target imaging distortion, add a mask to remove noise, and obtain the overall contour area of ​​the filtered billiard ball; it can identify the billiard ball outline and the size of the ball number; it can quickly calculate and identify the specific ball number of the target billiard ball in the billiard ball outline area; it can adapt to the intelligentization of billiard sports equipment systems and intelligent referee recognition, and can solve the problems of complex billiard sports laws, small targets, large numbers, and other characteristics in the recognition of multi-target ball sports. Rapid recognition of billiard targets and ball numbers.

[0044] The billiard target rapid identification method and system of the present invention, and other advantages, objectives and features of the present invention will be partially reflected through the following description, and will also be partially understood by technicians in this field through research and practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0046] Figure 1 This is a diagram showing an embodiment of the method and system for quickly identifying billiard targets of the present invention.

[0047] Figure 2 This is a diagram of an application example of the method and system for quickly identifying billiard targets of the present invention.

[0048] Figure 3 This is another application example of the method and system for quickly identifying billiard targets of the present invention. DETAILED DESCRIPTION

[0049] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments, so that those skilled in the art can implement the invention with reference to the instructions. As shown in the drawings, the present invention provides a method for quickly identifying a billiard target, comprising:

[0050] S100 uses a dual camera group to obtain multi-angle images of the target billiard ball, quickly crops the image and performs the first color gamut conversion through the machine vision platform, and performs elliptical distortion correction to obtain a circular correction image of the billiard ball imaging elliptical distortion;

[0051] S200, adding a mask to the circularly corrected image of the billiard ball imaging elliptical distortion, performing a first binarization process and filtering noise points, and obtaining an overall contour area of ​​the noise-filtered billiard ball;

[0052] S300, performing a second color gamut conversion and a second binarization process in the overall contour area of ​​the filtered billiard ball, identifying the white area of ​​the billiard ball contour, filtering the noise points in the white area, and calculating and identifying the size of the ball number;

[0053] S400, based on the circular deflection correction image of the billiard ball imaging elliptical distortion and the white area parameters of the billiard ball outline, removes the white area and filters the noise, performs an average operation on the remaining color area of ​​the billiard ball outline, and quickly calculates and identifies the specific ball number of the target billiard ball.

[0054] The principle and effect of the above technical solution are as follows: the present invention provides a method for rapid identification of billiard targets, including: obtaining multi-angle images of target billiards through a dual camera group, rapidly cropping images and performing the first color gamut conversion through a machine vision platform, and performing elliptical distortion correction to obtain a circular correction image of the elliptical distortion of billiard imaging; adding a mask to the circular correction image of the elliptical distortion of billiard imaging, performing the first binarization processing and filtering noise points, and obtaining the overall contour area of ​​the filtered billiard ball; performing the second color gamut conversion and the second binarization processing in the overall contour area of ​​the filtered billiard ball, identifying the white area of ​​the billiard outline, filtering the noise points in the white area, and calculating the size of the identified ball number; according to the circular correction image of the elliptical distortion of billiard imaging and the parameters of the white area of ​​the billiard outline, removing the white area and filtering the noise points, and calculating the remaining billiard outline. The color area is averaged, and the specific ball number of the target billiard ball is quickly calculated and identified; the multi-angle image of the target billiard ball can be obtained and the distortion of the target imaging and the cutting conversion speed can be improved; the distortion of the target imaging can be corrected, and a mask can be added to remove noise to obtain the overall contour area of ​​the filtered billiard ball; the outline of the billiard ball and the size of the ball number can be identified; the specific ball number of the target billiard ball can be quickly calculated and identified for the billiard ball contour area; it can adapt to the intelligentization of billiard sports equipment system and intelligent referee identification, and can solve the problem of complex billiard sports laws, small targets, large numbers and other characteristics of billiard targets and ball numbers in multi-target ball sports target identification; the following steps 1-12 are the program logic of the billiard target rapid identification and color selection algorithm of the present invention, and the program processes the image according to the execution order of 1234-abcd:

[0055] 1. Image cropping size: It accelerates recognition and reduces the amount of calculation;

[0056] 2. Scaling to maintain a fixed size: The efficiency of cropping size modification is significantly improved, and the complexity of subsequent algorithm parameters is significantly reduced;

[0057] 3. Gaussian blur processing: Select the Gaussian blur kernel parameters for image processing to reduce the noise and brightness changes of the image;

[0058] 4. Ball outline recognition:

[0059] a. Add mask: set the shape and size of the mask to cover irrelevant areas in the image, reduce the amount of calculation and completely eliminate the image information in irrelevant areas; b. Color gamut conversion: convert the original BGR color gamut into LAB and HSV color gamuts. These two color gamuts can better binarize the image in order to distinguish the outline of the ball. By fusing and comparing the results of the LAB and HSV color gamuts, the outline can be better obtained; c. Binarized image: empirically set the upper and lower limits of LAB and HSV corresponding to the optical definition, and use this value to binarize the image, and the color image becomes a black and white image; d. Edge detection: perform edge detection on the binary image to find all the contours in the image; e. Contour filtering: filter and filter all contours obtained based on edge detection, and set the corresponding minimum threshold based on the contour closure, contour length, and contour area. Contours smaller than this value are judged as noise points and deleted. In this way, the final outline of the ball is obtained; f. Image fusion: fuse the filtered contour with the original image to obtain an image with pixels only in the ball area;

[0060] 5. White contour recognition: Recognize the white contour on the ball and obtain a binary image to determine the size: a. Color gamut conversion: Convert the image (BGR) in 4.f to LAB and HSV color gamut again; these two color gamuts can better binarize the image to distinguish the white area of ​​the ball; b. Binarized image: Empirically set new LAB and HSV corresponding to the upper and lower limits of the optical definition, and use this value to binarize the image, and the color image becomes a black and white image; c. Edge detection: Perform edge detection on the binary image to find all white contours in the image; d. Contour filtering: Filter and filter all contours obtained based on edge detection, and set the corresponding minimum threshold through contour closure, contour length, and contour area. Contours smaller than this value are judged as noise points and deleted. In this way, the white contour of the final ball is obtained; e. Identify the size: Use calculation parameters, including: the number of white areas, area, arc length, approximate rectangular aspect ratio, and perimeter area ratio; to determine the size of the ball;

[0061] 6. Recognition of color contours: Recognition of color contours on the ball, obtaining a binary image, which is used to determine the ball number:

[0062] a. Color gamut conversion: Convert the image (BGR) in 4.f to LAB and HSV again. These two color gamuts can better binarize the image so as to distinguish the white area of ​​the ball;

[0063] b. Binarization of images: empirically set new LAB and HSV values ​​corresponding to the upper and lower limits of optical definitions, and use these values ​​to binarize the image, turning the color image into a black and white image;

[0064] c. Edge detection: Perform edge detection on the binary image to find all white contours in the image;

[0065] d. Contour filtering: Filter all contours obtained based on edge detection, and set the corresponding minimum threshold based on contour closure, contour length, and contour area. Contours smaller than this value are considered noise points and deleted. This will obtain the contour of the colored area of ​​the final ball.

[0066] 7. Image fusion: Based on the binary images and corresponding contours obtained in 4, 5, and 6, perform debinarization and image bit operations to obtain an image in which only the color area retains the original color and the other areas are all black;

[0067] 8. Color calculation: Perform edge detection based on the image in 7 to obtain the color contour position information. Perform mean calculation on the color within the contour range to obtain the color average of the ball;

[0068] 9. Color prediction: The color average value obtained in 8 is used to increase the elastic range size; the color average value corresponds to the LAB and HSV values; the elastic range size increase includes: weakening the recognition requirements, reducing the inability to judge the color due to the color average value not being in the database due to changes in light or angle; then comparing the color range values ​​of the balls learned in advance in the database, when the average value falls into the range value in the database after the elastic increase, it indicates that the ball may be a certain number in the database;

[0069] 10. Ball number judgment: Based on the results of the size and color prediction, the final result of the ball is logically judged; if it is not possible to make a clear judgment, the most likely ball number and the truth of the result are returned; true means the result is correct, false means the result is uncertain;

[0070] 11. Final ball number determination: Since two cameras are used, each ball will provide two results, including the ball number and the authenticity of the result. Based on the above results, the most likely ball number is determined by final logical judgment. If the recognition results of the two pictures are consistent, the result is output; if they are inconsistent, the final result is determined by logical judgment.

[0071] 12. About the study of the color range of the ball in the database: Through the above algorithm, the mean of the color is obtained. The mean values ​​obtained from multiple angles of the ball are taken as the color range with the maximum and minimum values. Due to the different camera installation positions and different lighting environments, the parameters of the above 1-12 steps algorithm are calculated using two sets.

[0072] In one embodiment, S100 includes:

[0073] S101, using a dual camera group to capture multi-angle images of target billiard balls, and quickly cropping the multi-angle images of the target billiard balls through a machine vision platform to obtain a cropped matrix area image;

[0074] S102, performing a first color gamut conversion on the cropped matrix region image to obtain a matrix region color gamut conversion image;

[0075] S103, according to the matrix area color gamut conversion image, correcting the elliptical distortion of the billiard image in the matrix area color gamut conversion image, and obtaining a circular correction image of the elliptical distortion of the billiard image.

[0076] The principle and effect of the above technical solution are as follows: using a dual camera group to shoot a target billiard ball multi-angle image of multiple target billiard balls, quickly cutting the target billiard ball multi-angle image through a machine vision platform, and obtaining a cut matrix area image; performing a first color gamut conversion on the cut matrix area image to obtain a matrix area color gamut conversion image; according to the matrix area color gamut conversion image, correcting the elliptical distortion matrix area of ​​the billiard ball image in the matrix area color gamut conversion image to obtain a billiard ball imaging elliptical distortion circular correction image; quickly cutting the target billiard ball multi-angle image through a machine vision platform, and obtaining the cut matrix area image includes: the machine vision platform according to The multi-angle image of the target billiard ball first identifies the outline of the billiard ball in the area where the billiard ball color and the background color do not overlap. For the area where the billiard ball color and the background color overlap, the circular arc of the non-overlapping area is extended and filled to form an extended and filled complete circle, and the extended and filled complete circular area is quickly cropped to obtain the cropped matrix area image; according to the matrix area color gamut conversion image, the elliptical distortion matrix area of ​​the billiard ball image in the matrix area color gamut conversion image is elliptical corrected to obtain the billiard ball imaging elliptical distortion circular correction image, including: respectively detecting and calculating the distance between the first camera imaging point of the dual camera group and the first target billiard ball, and obtaining the first imaging distance of the first target billiard ball; detecting and calculating The first imaging angle of the first target billiard ball is obtained by measuring the angle between the central optical axis of the first camera and the center point of the first target billiard ball; the distance between the imaging point of the second camera and the first target billiard ball is detected and calculated to obtain the second imaging distance of the first target billiard ball; the angle between the central optical axis of the second camera and the center point of the first target billiard ball is detected and calculated, and the second imaging angle of the first target billiard ball; the first eccentricity of the elliptical distortion of the image of the first target billiard ball is calculated according to the first imaging distance of the first target billiard ball and the first imaging angle of the first target billiard ball; the second eccentricity of the elliptical distortion of the image of the first target billiard ball is calculated according to the second imaging distance of the first target billiard ball and the second imaging angle of the first target billiard ball; the second eccentricity of the elliptical distortion of the image of the first target billiard ball is calculated according to the first eccentricity of the elliptical distortion , calculate a first adjustment ratio for perfect circle restoration; calculate a second adjustment ratio for perfect circle restoration according to the second eccentricity of the elliptical distortion; adjust the elliptical distortion of the first target billiard image according to the ratio value interval formed by the first adjustment ratio for perfect circle restoration and the second adjustment ratio for perfect circle restoration, until the elliptical distortion adjustment of the first target billiard image reaches the set perfect circle restoration degree, so that the elliptical distortion of the first target billiard image is restored to the billiard image perfect circle, and the first target billiard imaging elliptical distortion circular correction is performed; the second target billiard imaging elliptical distortion circular correction is performed in sequence until the elliptical distortion of all target billiard imaging is corrected, and the billiard imaging elliptical distortion circular correction image is obtained.

[0077] In one embodiment, S200 includes:

[0078] S201, adding a mask to the billiard ball imaging elliptical distortion circular correction image to cover the non-billiard ball area, and obtaining a billiard ball color gamut conversion mask image;

[0079] S202, by performing a first binarization process on the billiard ball color gamut conversion mask image, the overall outline of the billiard ball is identified, and noise points are filtered to obtain a noise-filtered billiard ball overall outline region.

[0080] The principle and effect of the above technical solution are as follows: adding a mask to the circularly corrected image of the billiard ball imaging elliptical distortion to cover the non-billiard ball area and obtain the billiard ball color gamut conversion mask image; by performing the first binarization processing on the billiard ball color gamut conversion mask image, the overall outline of the billiard ball is identified, and the noise is filtered to obtain the overall outline area of ​​the noise-filtered billiard ball.

[0081] In one embodiment, S300 includes:

[0082] S301, performing a second color gamut conversion in the noise-filtered billiard ball overall contour region to obtain a color gamut converted billiard ball contour region;

[0083] S302, converting the billiard ball outline area according to the color gamut, performing a second binarization process, identifying the white area of ​​the billiard ball outline and filtering the noise in the white area, and calculating and identifying the size of the ball number;

[0084] The parameters of the white area of ​​the billiard ball outline include: the shape of the white area of ​​the billiard ball outline, the perimeter of the white area of ​​the billiard ball outline and the area of ​​the white area of ​​the billiard ball outline.

[0085] The principle and effect of the above technical solution are as follows: performing a second color gamut conversion in the overall contour area of ​​the noise-filtered billiard ball to obtain the color gamut converted billiard ball contour area; performing a second binarization process according to the color gamut converted billiard ball contour area to identify the white area of ​​the billiard ball contour and filter the noise in the white area, and calculate the size of the identified ball number; the parameters of the white area of ​​the billiard ball contour include: the shape of the white area of ​​the billiard ball contour, the perimeter of the white area of ​​the billiard ball contour and the area of ​​the white area of ​​the billiard ball contour; performing a second binarization process according to the color gamut converted billiard ball contour area to identify the white area of ​​the billiard ball contour and filter the noise in the white area, and calculating the size of the identified ball number includes: performing a second binarization process according to the color gamut converted billiard ball contour area to identify the white area of ​​the billiard ball contour; filtering the noise in the white area according to the parameters of the white area of ​​the billiard ball contour, and retaining the white area after the noise is filtered; and calculating the size of the identified ball number according to the noise filtering white area.

[0086] In one embodiment, S400 includes:

[0087] S401, according to the circular deflection correction image of the billiard ball imaging elliptical distortion and the white area parameters of the billiard ball outline, the white area is removed and the noise is filtered, and only the billiard ball outline area is retained to obtain the remaining color area of ​​the billiard ball outline;

[0088] S402, performing a mean value operation on the remaining color area of ​​the billiard ball outline to obtain a remaining color average value; and quickly calculating and identifying a specific ball number of the target billiard ball based on the remaining color average value.

[0089] The principle and effect of the above technical solution are as follows: according to the circular correction image of the elliptical distortion of the billiard ball imaging and the parameters of the white area of ​​the billiard ball outline, the white area is removed and the noise is filtered, and only the billiard ball outline area is retained to obtain the residual color area of ​​the billiard ball outline; the residual color area of ​​the billiard ball outline is averaged to obtain the residual color average value; the specific ball number of the target billiard ball can be quickly calculated and identified according to the residual color average value; the specific ball number of the target billiard ball can be quickly calculated and identified for the billiard ball outline area; it can adapt to the intelligentization of billiard sports equipment system and intelligent referee identification, and can solve the problem of rapid identification of billiard targets and ball numbers with complex billiard sports laws, small targets, large numbers and other characteristics in the target identification of multi-target ball sports;

[0090] Make logical judgment based on the color sorting results of the dual cameras:

[0091] Because the angles of the table tennis balls when they reach the camera are different and the reflectivity of the surfaces of table tennis balls of different brands is different, the results of a single camera are prone to errors, especially the judgment of their colors. Therefore, the accuracy of color sorting and recognition can be significantly improved by shooting with dual cameras for detection and recognition, and then performing logical judgment on the detection and recognition results;

[0092] The recognition result of a single camera includes the following information: 1. Large or small ball, the recognition result of large or small ball is unique, which uniquely reflects whether the ball is large or small this time; 2. The recognition result of the ball number is not accurate. For example, for a red ball No. 11, if the white part is facing the camera in a large area, and the colored area is only exposed in a small circle on the outside, due to different lighting conditions, the average color value may fall into the recognition range of multiple balls; the final color recognition result may be No. 3, 11, 12, 13; 3. The judgment result of the final result output; 4. The true value reflects the truthfulness of the recognition result;

[0093] Single camera logic judgment:

[0094] Scenario 1: The ball number and color are uniquely identified, and the final result is presented based on the identification of the size; for example:

[0095] Ball number and color: 3; Size: small; Final result: ball No. 3; True value: true;

[0096] Ball number color: 3; Size: Large; Final result: Ball No. 11; True value: True;

[0097] Similarly, if the color is 11, the color of the 11 ball and the 3 ball is the same, and the final result changes accordingly according to the size;

[0098] Case 2: The ball number and color are not unique, but the colors of the same size are in the result;

[0099] Ball number color: 3, 11, 12; Size: Small; Final result: Ball No. 3; True value: True;

[0100] Ball number color: 3, 11, 12; Size: Large; Final result: Ball No. 11; True value: True;

[0101] Because 3 and 11 are the same color, according to the recognition statistics analysis, when the same color appears in a group, the accuracy of the result is very high;

[0102] Case 3: The ball number color is not unique, and no color with the same size appears in the result;

[0103] Color: 11, 12, 13; Size: Small; Final result: Small ball; True value: False;

[0104] Color: 11, 12, 13; Size: Large; Final result: Large ball; True value: False;

[0105] Case 4: No color recognition result:

[0106] Color: None; Size: Small; Final Result: Small Ball; True Value: False;

[0107] Color: None; Size: Large; Final Result: Large Ball; True Value: False;

[0108] The above is the judgment logic of the recognition result of a single camera;

[0109] Perform logical judgment based on the results of the dual cameras. The logical judgment of the dual camera results includes:

[0110] Due to the actual installation situation, the recognition result of camera B is assumed to be more accurate. Therefore, when the results of cameras A and B are inconsistent, the result of camera B is preferred.

[0111] Case 1: If the results of cameras AB are consistent, the result is directly output, and the truth value of the result is true;

[0112] Case 2: AB results are both small or large balls, the truth value of this result is false;

[0113] Camera A: color result; Camera B: color result; Final result: color of Camera B;

[0114] Camera A: no color result; Camera B: with color result; Final result: the color of Camera B;

[0115] Camera A: with color result; Camera B: without color result; Final result: the color of Camera A;

[0116] Scenario 3: The size of the results of AB is inconsistent, and the size of the result of camera B is output. The color is judged as follows, and the true value of the result is false;

[0117] Camera A: color result; Camera B: color result; Final result: color of Camera B;

[0118] Camera A: no color result; Camera B: with color result; Final result: the color of Camera B;

[0119] Camera A: with color result; Camera B: without color result; Final result: the color of Camera A;

[0120] Logical judgment at the ball placement end: If the logical judgment value of the dual cameras is true, the ball is placed in the ball position that needs to ensure the accuracy of the result first; otherwise, the ball is placed in the uncritical ball position in order;

[0121] The recognition results of the dual-camera group significantly improve the recognition efficiency and accuracy in various situations, such as the recognition process of a large number of ball rolling.

[0122] The present invention provides a billiard target rapid identification system, comprising:

[0123] The module for fast cutting and correction of elliptical distortion acquires multi-angle images of the target billiard ball through a dual-camera group, quickly cuts the image and performs the first color gamut conversion through a machine vision platform, and performs elliptical distortion correction to obtain a circular correction image of the elliptical distortion of the billiard ball imaging;

[0124] The mask processing noise filtering module adds a mask to the billiard ball imaging elliptical distortion circular correction image, performs the first binarization processing and filters the noise to obtain the overall contour area of ​​the noise-filtered billiard ball;

[0125] The color gamut conversion contour size recognition module performs a second color gamut conversion and a second binarization process in the overall contour area of ​​the filtered noise billiard ball, identifies the white area of ​​the billiard ball contour, filters the noise in the white area, and calculates and recognizes the size of the ball number;

[0126] The regional mean ball number recognition module removes the white area and filters the noise based on the circular correction image of the billiard ball's elliptical distortion and the white area parameters of the billiard ball's outline, performs a mean operation on the remaining color area of ​​the billiard ball's outline, and quickly calculates and identifies the specific ball number of the target billiard ball.

[0127] The principle and effect of the above technical solution are as follows: the present invention provides a billiard target rapid identification system, including: a rapid cropping elliptical distortion correction module 100, which obtains multi-angle images of the target billiard ball through a dual camera group, rapidly crops the image and performs the first color gamut conversion through a machine vision platform, and performs elliptical distortion correction to obtain a billiard imaging elliptical distortion circular correction image; a mask processing noise filtering module 200, which adds a mask to the billiard imaging elliptical distortion circular correction image, performs the first binarization processing and filters the noise, and obtains the overall contour area of ​​the filtered billiard ball; a color gamut conversion contour size recognition module 300, which performs the second color gamut conversion and the second binarization processing in the overall contour area of ​​the filtered billiard ball, identifies the white area of ​​the billiard ball contour, filters the noise in the white area, and calculates and identifies the ball number; a regional mean ball number recognition module 400, which calculates and identifies the ball number according to the billiard imaging elliptical distortion circular correction image and The white area parameters of the billiard ball outline are removed, the white area is removed and the noise is filtered, and the remaining color area of ​​the billiard ball outline is averaged to quickly calculate and identify the specific ball number of the target billiard ball; the multi-angle image of the target billiard ball can be obtained and the distortion of the target imaging and the cutting conversion speed can be improved; the distortion of the target imaging can be corrected, and a mask can be added to remove the noise to obtain the overall contour area of ​​the filtered billiard ball; the billiard ball outline and the size of the ball number can be identified; the specific ball number of the target billiard ball can be quickly calculated and identified for the billiard ball contour area; it can adapt to the intelligentization of billiard sports equipment system and intelligent referee identification, and can solve the billiard ball target and ball number rapid identification with complex billiard sports laws, small targets, and large numbers in the target identification of multi-target ball sports; the following steps 1-12 are the program logic of the billiard ball target rapid identification and color selection algorithm of the present invention, and the program processes the image according to the execution order of 1234-abcd:

[0128] 1. Image cropping size: It accelerates recognition and reduces the amount of calculation;

[0129] 2. Scaling to maintain a fixed size: The efficiency of cropping size modification is significantly improved, and the complexity of subsequent algorithm parameters is significantly reduced;

[0130] 3. Gaussian blur processing: Select the Gaussian blur kernel parameters for image processing to reduce the noise and brightness changes of the image;

[0131] 4. Ball outline recognition:

[0132] a. Add mask: set the shape and size of the mask to cover irrelevant areas in the image, reduce the amount of calculation and completely eliminate the image information in irrelevant areas; b. Color gamut conversion: convert the original BGR color gamut into LAB and HSV color gamuts. These two color gamuts can better binarize the image in order to distinguish the outline of the ball. By fusing and comparing the results of the LAB and HSV color gamuts, the outline can be better obtained; c. Binarized image: empirically set the upper and lower limits of LAB and HSV corresponding to the optical definition, and use this value to binarize the image, and the color image becomes a black and white image; d. Edge detection: perform edge detection on the binary image to find all the contours in the image; e. Contour filtering: filter and filter all contours obtained based on edge detection, and set the corresponding minimum threshold based on the contour closure, contour length, and contour area. Contours smaller than this value are judged as noise points and deleted. In this way, the final outline of the ball is obtained; f. Image fusion: fuse the filtered contour with the original image to obtain an image with pixels only in the ball area;

[0133] 5. White contour recognition: Recognize the white contour on the ball and obtain a binary image to determine the size: a. Color gamut conversion: Convert the image (BGR) in 4.f to LAB and HSV color gamut again; these two color gamuts can better binarize the image to distinguish the white area of ​​the ball; b. Binarized image: Empirically set new LAB and HSV corresponding to the upper and lower limits of the optical definition, and use this value to binarize the image, and the color image becomes a black and white image; c. Edge detection: Perform edge detection on the binary image to find all white contours in the image; d. Contour filtering: Filter and filter all contours obtained based on edge detection, and set the corresponding minimum threshold through contour closure, contour length, and contour area. Contours smaller than this value are judged as noise points and deleted. In this way, the white contour of the final ball is obtained; e. Identify the size: Use calculation parameters, including: the number of white areas, area, arc length, approximate rectangular aspect ratio, and perimeter area ratio; to determine the size of the ball;

[0134] 6. Recognition of color contours: Recognition of color contours on the ball, obtaining a binary image, which is used to determine the ball number:

[0135] a. Color gamut conversion: Convert the image (BGR) in 4.f to LAB and HSV again. These two color gamuts can better binarize the image so as to distinguish the white area of ​​the ball;

[0136] b. Binarization of images: empirically set new LAB and HSV values ​​corresponding to the upper and lower limits of optical definitions, and use these values ​​to binarize the image, turning the color image into a black and white image;

[0137] c. Edge detection: Perform edge detection on the binary image to find all white contours in the image;

[0138] d. Contour filtering: Filter all contours obtained based on edge detection, and set the corresponding minimum threshold based on contour closure, contour length, and contour area. Contours smaller than this value are considered noise points and deleted. This will obtain the contour of the colored area of ​​the final ball.

[0139] 7. Image fusion: Based on the binary images and corresponding contours obtained in 4, 5, and 6, perform debinarization and image bit operations to obtain an image in which only the color area retains the original color and the other areas are all black;

[0140] 8. Color calculation: Perform edge detection based on the image in 7 to obtain the color contour position information. Perform mean calculation on the color within the contour range to obtain the color average of the ball;

[0141] 9. Color prediction: The color average value obtained in 8 is used to increase the elastic range size; the color average value corresponds to the LAB and HSV values; the elastic range size increase includes: weakening the recognition requirements, reducing the inability to judge the color due to the color average value not being in the database due to changes in light or angle; then comparing the color range values ​​of the balls learned in advance in the database, when the average value falls into the range value in the database after the elastic increase, it indicates that the ball may be a certain number in the database;

[0142] 10. Ball number judgment: Based on the results of the size and color prediction, the final result of the ball is logically judged; if it is not possible to make a clear judgment, the most likely ball number and the truth of the result are returned; true means the result is correct, false means the result is uncertain;

[0143] 11. Final ball number determination: Since two cameras are used, each ball will provide two results, including the ball number and the authenticity of the result. Based on the above results, the most likely ball number is determined by final logical judgment. If the recognition results of the two pictures are consistent, the result is output; if they are inconsistent, the final result is determined by logical judgment.

[0144] 12. About the study of the color range of the ball in the database: Through the above algorithm, the mean of the color is obtained. The mean values ​​obtained from multiple angles of the ball are taken as the color range with the maximum and minimum values. Due to the different camera installation positions and different lighting environments, the parameters of the above 1-12 steps algorithm are calculated using two sets.

[0145] In one embodiment, a fast cropping ellipse distortion correction module includes:

[0146] The machine vision fast cutting unit uses a dual camera group to capture multi-angle images of target billiard balls, and quickly cuts the multi-angle images of target billiard balls through a machine vision platform to obtain the cut matrix area image;

[0147] A matrix area color gamut conversion unit performs a first color gamut conversion on the cropped matrix area image to obtain a matrix area color gamut conversion image;

[0148] The billiard imaging elliptical distortion correction unit performs elliptical distortion correction on the billiard image elliptical distortion matrix area in the matrix area gamut conversion image according to the matrix area gamut conversion image, and obtains the billiard imaging elliptical distortion circular correction image.

[0149] The principle and effect of the above technical solution are: a fast cutting elliptical distortion correction module, including: a machine vision fast cutting unit, which uses a dual camera group to shoot a target billiard ball multi-angle image of multiple target billiard balls, and quickly cuts the target billiard ball multi-angle image through a machine vision platform to obtain a cut matrix area image; a matrix area color gamut conversion unit, which performs a first color gamut conversion on the cut matrix area image to obtain a matrix area color gamut conversion image; a billiard imaging elliptical correction unit, which corrects the elliptical distortion matrix area of ​​the billiard image in the matrix area color gamut conversion image according to the matrix area color gamut conversion image, and obtains a billiard imaging elliptical distortion circular correction image; the target billiard ball multi-angle The image is quickly cropped through a machine vision platform, and the matrix area image after cropping is obtained, including: the machine vision platform first identifies the billiard ball outline in the area where the billiard ball color and the background color do not overlap according to the multi-angle image of the target billiard ball, and for the area where the billiard ball color and the background color overlap, the circular arc of the non-overlapping area is extended and filled to form an extended and filled complete circle, and the extended and filled complete circle area is quickly cropped to obtain the matrix area image after cropping; according to the matrix area color gamut conversion image, the elliptical distortion matrix area of ​​the billiard ball image in the matrix area color gamut conversion image is elliptical corrected to obtain the billiard ball imaging elliptical distortion circular correction image, including: respectively detecting and calculating the imaging point of the first camera of the dual camera group and the first camera The distance between the first target billiard ball and the first target billiard ball is detected and calculated, and the first imaging distance of the first target billiard ball is obtained; the angle between the central optical axis of the first camera and the center point of the first target billiard ball is detected and calculated, and the first imaging angle of the first target billiard ball is obtained; the distance between the imaging point of the second camera and the first target billiard ball is detected and calculated, and the second imaging angle of the first target billiard ball is obtained; according to the first imaging distance of the first target billiard ball and the first imaging angle of the first target billiard ball, the first eccentricity of the elliptical distortion of the image of the first target billiard ball is calculated; according to the second imaging distance of the first target billiard ball and the second imaging angle of the first target billiard ball, the second eccentricity of the elliptical distortion of the image of the first target billiard ball is calculated; According to the first eccentricity of the elliptical distortion, a first adjustment ratio for perfect circle restoration is calculated; according to the second eccentricity of the elliptical distortion, a second adjustment ratio for perfect circle restoration is calculated; according to the ratio value interval formed by the first adjustment ratio for perfect circle restoration and the second adjustment ratio for perfect circle restoration, the elliptical distortion of the first target billiard image is adjusted until the elliptical distortion adjustment of the first target billiard image reaches the set perfect circle restoration degree, so that the elliptical distortion of the first target billiard image is restored to the billiard image perfect circle, and the elliptical distortion of the first target billiard image is corrected circularly; the elliptical distortion of the second target billiard image is corrected circularly until the elliptical distortion of all target billiard images is corrected circularly, and the elliptical distortion of the billiard image is obtained.

[0150] In one embodiment, the mask processing noise filtering module includes:

[0151] The mask adding area covering unit adds a mask to the billiard imaging elliptical distortion circular correction image to cover the non-billiard area and obtain the billiard color gamut conversion mask image;

[0152] The contour recognition noise filtering unit performs the first binarization processing on the billiard ball color gamut conversion mask image, recognizes the overall contour of the billiard ball, filters the noise, and obtains the overall contour area of ​​the noise-filtered billiard ball.

[0153] The principle and effect of the above technical solution are: a mask processing noise filtering module, including: a mask adding area covering unit, which adds a mask to the billiard ball imaging elliptical distortion circular correction image to cover the non-billiard ball area and obtain the billiard ball color gamut conversion mask image; a contour recognition noise filtering unit, which identifies the overall contour of the billiard ball by performing the first binarization processing on the billiard ball color gamut conversion mask image, and filters the noise to obtain the overall contour area of ​​the filtered billiard ball.

[0154] In one embodiment, the color gamut conversion contour size recognition module includes:

[0155] The contour area color gamut conversion unit performs a second color gamut conversion in the overall contour area of ​​the noise-filtered billiard ball to obtain a color gamut converted billiard ball contour area;

[0156] The conversion area white recognition unit converts the billiard ball outline area according to the color gamut, performs a second binarization process, identifies the white area of ​​the billiard ball outline and filters the noise in the white area, and calculates and recognizes the size of the ball number;

[0157] The parameters of the white area of ​​the billiard ball outline include: the shape of the white area of ​​the billiard ball outline, the perimeter of the white area of ​​the billiard ball outline and the area of ​​the white area of ​​the billiard ball outline.

[0158] The principle and effect of the above technical solution are as follows: a color gamut conversion contour size recognition module includes: a contour area color gamut conversion unit, which performs a second color gamut conversion in the overall contour area of ​​the noise-filtered billiard ball to obtain a color gamut converted billiard ball contour area; a conversion area white recognition unit, which performs a second binarization process according to the color gamut converted billiard ball contour area, identifies the white area of ​​the billiard ball contour and filters the noise in the white area, and calculates and identifies the size of the ball; the parameters of the white area of ​​the billiard ball contour include: the shape of the white area of ​​the billiard ball contour, the perimeter of the white area of ​​the billiard ball contour and the area of ​​the white area of ​​the billiard ball contour; the billiard ball contour area is converted according to the color gamut, and the second binarization process is performed to identify the white area of ​​the billiard ball contour and filter the noise in the white area, and the calculation of the size of the identified ball number includes: the billiard ball contour area is converted according to the color gamut, and the second binarization process is performed to identify the white area of ​​the billiard ball contour; the white area noise is filtered according to the parameters of the billiard ball contour, and the white area after the noise is filtered is retained; the white area is filtered according to the noise, and the size of the identified ball number is calculated.

[0159] In one embodiment, the regional average ball number recognition module includes:

[0160] The residual color area filtering unit removes the white area and filters the noise according to the circular correction image of the billiard ball imaging elliptical distortion and the white area parameters of the billiard ball outline, and only retains the billiard ball outline area to obtain the residual color area of ​​the billiard ball outline;

[0161] The mean operation ball number fast identification unit performs a mean operation on the remaining color area of ​​the billiard ball outline to obtain the remaining color average value; based on the remaining color average value, the specific ball number of the target billiard ball is quickly calculated and identified.

[0162] The principle and effect of the above technical solution are as follows: the regional mean ball number recognition module includes: a residual color area filtering unit, which removes the white area and filters the noise according to the billiard ball imaging elliptical distortion circular correction image and the billiard ball outline white area parameters, retains only the billiard ball outline area, and obtains the billiard ball outline residual color area; a mean operation ball number fast recognition unit, which performs a mean operation on the billiard ball outline residual color area to obtain the residual color average value; the specific ball number of the target billiard ball can be quickly calculated and identified according to the residual color average value; the specific ball number of the target billiard ball can be quickly calculated and identified for the billiard ball outline area; it can adapt to the intelligentization of billiard sports equipment system and intelligent referee recognition, and can solve the billiard ball target and ball number fast recognition with complex billiard sports laws, small targets, large numbers and other characteristics in multi-target ball sports target recognition;

[0163] Make logical judgment based on the color sorting results of the dual cameras:

[0164] Because the angles of the table tennis balls when they reach the camera are different and the reflectivity of the surfaces of table tennis balls of different brands is different, the results of a single camera are prone to errors, especially the judgment of their colors. Therefore, the accuracy of color sorting and recognition can be significantly improved by shooting with dual cameras for detection and recognition, and then performing logical judgment on the detection and recognition results;

[0165] The recognition results of a single camera include the following information:

[0166] 1. Large or small ball, the recognition result of large or small ball is unique, which uniquely reflects whether the ball is large or small this time; 2. The ball number recognition result is not accurate. For example, for a red ball No. 11, if the white part is facing the camera in a large area, and the colored area is only exposed in a small circle on the outside, due to different lighting conditions, its color average value may fall into the recognition range of multiple balls; the final color recognition result may be No. 3, 11, 12, 13; 3. The judgment result of the final result output; 4. The true value reflects the truthfulness of the recognition result;

[0167] Single camera logic judgment:

[0168] Scenario 1: The ball number and color are uniquely identified, and the final result is presented based on the identification of the size; for example:

[0169] Ball number and color: 3; Size: small; Final result: ball No. 3; True value: true;

[0170] Ball number color: 3; Size: Large; Final result: Ball No. 11; True value: True;

[0171] Similarly, if the color is 11, the color of the 11 ball and the 3 ball is the same, and the final result changes accordingly according to the size;

[0172] Case 2: The ball number and color are not unique, but the colors of the same size are in the result;

[0173] Ball number color: 3, 11, 12; Size: Small; Final result: Ball No. 3; True value: True;

[0174] Ball number color: 3, 11, 12; Size: Large; Final result: Ball No. 11; True value: True;

[0175] Because 3 and 11 are the same color, according to the recognition statistics analysis, when the same color appears in a group, the accuracy of the result is very high;

[0176] Case 3: The ball number color is not unique, and no color with the same size appears in the result;

[0177] Color: 11, 12, 13; Size: Small; Final result: Small ball; True value: False;

[0178] Color: 11, 12, 13; Size: Large; Final result: Large ball; True value: False;

[0179] Case 4: No color recognition result:

[0180] Color: None; Size: Small; Final Result: Small Ball; True Value: False;

[0181] Color: None; Size: Large; Final Result: Large Ball; True Value: False;

[0182] The above is the judgment logic of the recognition result of a single camera;

[0183] Perform logical judgment based on the results of the dual cameras. The logical judgment of the dual camera results includes:

[0184] Due to the actual installation situation, the recognition result of camera B is assumed to be more accurate. Therefore, when the results of cameras A and B are inconsistent, the result of camera B is preferred.

[0185] Case 1: If the results of cameras AB are consistent, the result is directly output, and the truth value of the result is true;

[0186] Case 2: AB results are both small or large balls, the truth value of this result is false;

[0187] Camera A: color result; Camera B: color result; Final result: color of Camera B;

[0188] Camera A: no color result; Camera B: with color result; Final result: the color of Camera B;

[0189] Camera A: with color result; Camera B: without color result; Final result: the color of Camera A;

[0190] Scenario 3: The size of the results of AB is inconsistent, and the size of the result of camera B is output. The color is judged as follows, and the true value of the result is false;

[0191] Camera A: color result; Camera B: color result; Final result: color of Camera B;

[0192] Camera A: no color result; Camera B: with color result; Final result: the color of Camera B;

[0193] Camera A: with color result; Camera B: without color result; Final result: the color of Camera A;

[0194] Logical judgment at the ball placement end: If the logical judgment value of the dual cameras is true, the ball is placed in the ball position that needs to ensure the accuracy of the result first; otherwise, the ball is placed in the uncritical ball position in order;

[0195] The recognition results of the dual-camera group significantly improve the recognition efficiency and accuracy in various situations, such as the recognition process of a large number of ball rolling.

[0196] Although the embodiments of the present invention have been disclosed as above, they are not limited to the applications listed in the specification and the implementation methods. They can be fully applied to various fields suitable for the present invention. For those familiar with the art, additional modifications can be easily implemented. Therefore, without departing from the general concept defined by the claims and the scope of equivalents, the present invention is not limited to the specific details and the illustrations shown and described herein.

Claims

1. A method for quickly identifying billiard targets, characterized in that: include: S100 uses a dual camera group to obtain multi-angle images of the target billiard ball, quickly crops the image and performs the first color gamut conversion through the machine vision platform, and performs elliptical distortion correction to obtain a circular correction image of the billiard ball imaging elliptical distortion; S200, adding a mask to the circularly corrected image of the billiard ball imaging elliptical distortion, performing a first binarization process and filtering noise points, and obtaining an overall contour area of ​​the noise-filtered billiard ball; S300, performing a second color gamut conversion and a second binarization process in the overall contour area of ​​the filtered billiard ball, identifying the white area of ​​the billiard ball contour, filtering the noise points in the white area, and calculating and identifying the size of the ball number; S400, based on the circular deflection correction image of the billiard ball imaging elliptical distortion and the white area parameters of the billiard ball outline, removes the white area and filters the noise, performs an average operation on the remaining color area of ​​the billiard ball outline, and quickly calculates and identifies the specific ball number of the target billiard ball.

2. The method for quickly identifying a billiard ball target according to claim 1, characterized in that: S100 includes: S101, using a dual camera group to capture multi-angle images of target billiard balls, and quickly cropping the multi-angle images of the target billiard balls through a machine vision platform to obtain a cropped matrix area image; S102, performing a first color gamut conversion on the cropped matrix region image to obtain a matrix region color gamut conversion image; S103, according to the matrix area color gamut conversion image, correcting the elliptical distortion of the billiard image in the matrix area color gamut conversion image, and obtaining a circular correction image of the elliptical distortion of the billiard image.

3. The method for quickly identifying a billiard ball target according to claim 1, characterized in that: S200 includes: S201, adding a mask to the billiard ball imaging elliptical distortion circular correction image to cover the non-billiard ball area, and obtaining a billiard ball color gamut conversion mask image; S202, by performing a first binarization process on the billiard ball color gamut conversion mask image, the overall outline of the billiard ball is identified, and noise points are filtered to obtain a noise-filtered billiard ball overall outline region.

4. The method for quickly identifying a billiard ball target according to claim 1, characterized in that: S300 includes: S301, performing a second color gamut conversion in the noise-filtered billiard ball overall contour region to obtain a color gamut converted billiard ball contour region; S302, converting the billiard ball outline area according to the color gamut, performing a second binarization process, identifying the white area of ​​the billiard ball outline and filtering the noise in the white area, and calculating and identifying the size of the ball number; The parameters of the white area of ​​the billiard ball outline include: the shape of the white area of ​​the billiard ball outline, the perimeter of the white area of ​​the billiard ball outline and the area of ​​the white area of ​​the billiard ball outline.

5. The method for quickly identifying a billiard ball target according to claim 1, characterized in that: S400 includes: S401, according to the circular deflection correction image of the billiard ball imaging elliptical distortion and the white area parameters of the billiard ball outline, the white area is removed and the noise is filtered, and only the billiard ball outline area is retained to obtain the remaining color area of ​​the billiard ball outline; S402, performing a mean value operation on the remaining color area of ​​the billiard ball outline to obtain a remaining color average value; and quickly calculating and identifying a specific ball number of the target billiard ball based on the remaining color average value.

6. Billiards target rapid identification system, characterized in that: include: The module for fast cutting and correction of elliptical distortion acquires multi-angle images of the target billiard ball through a dual-camera group, quickly cuts the image and performs the first color gamut conversion through a machine vision platform, and performs elliptical distortion correction to obtain a circular correction image of the elliptical distortion of the billiard ball imaging; The mask processing noise filtering module adds a mask to the billiard ball imaging elliptical distortion circular correction image, performs the first binarization processing and filters the noise to obtain the overall contour area of ​​the noise-filtered billiard ball; The color gamut conversion contour size recognition module performs a second color gamut conversion and a second binarization process in the overall contour area of ​​the filtered noise billiard ball, identifies the white area of ​​the billiard ball contour, filters the noise in the white area, and calculates and recognizes the size of the ball number; The regional mean ball number recognition module removes the white area and filters the noise based on the circular correction image of the billiard ball's elliptical distortion and the white area parameters of the billiard ball's outline, performs a mean operation on the remaining color area of ​​the billiard ball's outline, and quickly calculates and identifies the specific ball number of the target billiard ball.

7. The billiard target rapid identification system according to claim 6, characterized in that: Fast cutting ellipse distortion correction module, including: The machine vision fast cutting unit uses a dual camera group to capture multi-angle images of target billiard balls, and quickly cuts the multi-angle images of target billiard balls through a machine vision platform to obtain the cut matrix area image; A matrix area color gamut conversion unit performs a first color gamut conversion on the cropped matrix area image to obtain a matrix area color gamut conversion image; The billiard imaging elliptical distortion correction unit performs elliptical distortion correction on the billiard image elliptical distortion matrix area in the matrix area gamut conversion image according to the matrix area gamut conversion image, and obtains the billiard imaging elliptical distortion circular correction image.

8. The billiard target rapid identification system according to claim 6, characterized in that: The mask processing noise filtering module includes: a mask adding area covering unit, which adds a mask to the billiard imaging elliptical distortion circular correction image to cover the non-billiard area and obtain the billiard color gamut conversion mask image; The contour recognition noise filtering unit performs the first binarization processing on the billiard ball color gamut conversion mask image, recognizes the overall contour of the billiard ball, filters the noise, and obtains the overall contour area of ​​the noise-filtered billiard ball.

9. The billiard target rapid identification system according to claim 6, characterized in that: Color gamut conversion contour size recognition module, including: The contour area color gamut conversion unit performs a second color gamut conversion in the overall contour area of ​​the noise-filtered billiard ball to obtain a color gamut converted billiard ball contour area; The conversion area white recognition unit converts the billiard ball outline area according to the color gamut, performs a second binarization process, identifies the white area of ​​the billiard ball outline and filters the noise in the white area, and calculates and recognizes the size of the ball number; The parameters of the white area of ​​the billiard ball outline include: the shape of the white area of ​​the billiard ball outline, the perimeter of the white area of ​​the billiard ball outline and the area of ​​the white area of ​​the billiard ball outline.

10. The billiard target rapid identification system according to claim 6, characterized in that: The regional mean ball number recognition module includes: The residual color area filtering unit removes the white area and filters the noise according to the circular correction image of the billiard ball imaging elliptical distortion and the white area parameters of the billiard ball outline, and only retains the billiard ball outline area to obtain the residual color area of ​​the billiard ball outline; The mean operation ball number fast identification unit performs a mean operation on the remaining color area of ​​the billiard ball outline to obtain the remaining color average value; based on the remaining color average value, the specific ball number of the target billiard ball is quickly calculated and identified.

Images