CN117314902A - Visual inspection method for flatness of basketball floor surface - Google Patents
Visual inspection method for flatness of basketball floor surface Download PDFInfo
- Publication number
- CN117314902A CN117314902A CN202311596693.0A CN202311596693A CN117314902A CN 117314902 A CN117314902 A CN 117314902A CN 202311596693 A CN202311596693 A CN 202311596693A CN 117314902 A CN117314902 A CN 117314902A
- Authority
- CN
- China
- Prior art keywords
- basketball floor
- area
- basketball
- floor image
- defect area
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims description 41
- 238000011179 visual inspection Methods 0.000 title claims description 16
- 230000007547 defect Effects 0.000 claims abstract description 139
- 230000002950 deficient Effects 0.000 claims abstract description 71
- 239000000203 mixture Substances 0.000 claims abstract description 28
- 238000001514 detection method Methods 0.000 claims abstract description 10
- 230000000007 visual effect Effects 0.000 claims abstract description 5
- 230000008859 change Effects 0.000 claims description 29
- 230000008569 process Effects 0.000 claims description 9
- 230000003595 spectral effect Effects 0.000 claims description 4
- 230000007423 decrease Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000005286 illumination Methods 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/66—Analysis of geometric attributes of image moments or centre of gravity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
- G06V10/507—Summing image-intensity values; Histogram projection analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
- G06V10/758—Involving statistics of pixels or of feature values, e.g. histogram matching
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/762—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using clustering, e.g. of similar faces in social networks
- G06V10/763—Non-hierarchical techniques, e.g. based on statistics of modelling distributions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
- G06V10/765—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects using rules for classification or partitioning the feature space
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Geometry (AREA)
- Quality & Reliability (AREA)
- Probability & Statistics with Applications (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention relates to the technical field of image processing, in particular to a visual detection method for the flatness of the surface of a basketball floor, which comprises the following steps: acquiring a Gaussian mixture model according to a gray histogram of the basketball floor image, and acquiring a gray value range of a normal area according to a mean value parameter and a standard deviation parameter of the Gaussian model in the Gaussian mixture model so as to obtain a defective pixel point; acquiring each adjacent node pair according to the defective pixel point; according to each adjacent node pair, the offset degree between each adjacent node pair is obtained, a weighted undirected graph is further obtained, each initial category of the basketball floor image is obtained according to the weighted undirected graph, the circle center of each fitting circle is further obtained, and each final category of the basketball floor image is obtained according to the circle center of each fitting circle; acquiring each defect area according to each final category of the basketball floor image; the characteristic value of each defect area is obtained, and then the type of each defect area is judged.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to a visual detection method for the flatness of the surface of a basketball floor.
Background
The flatness detection of the basketball court floor surface can ensure the fairness and the safety of the game, and if the floor is uneven, unsafe conditions such as falling and the like can occur when players move and dribble. Secondly, uneven basketball floors can influence the rolling track of the basketball, so that the technical problem in the competition process is caused. Therefore, the detection of the flatness of the basketball floor surface is very important, namely, the defect area of the basketball floor surface is detected, so that the staff can conveniently use different treatment methods according to the types of different defect areas.
The defect areas such as concave or convex areas exist on the surface of the basketball floor, and the gray features expressed by the illumination on different types of defects are different, so that the complete defect areas cannot be accurately segmented only by the existing segmentation algorithm, and the subsequent process of distinguishing the types of the defect areas is inaccurate.
Disclosure of Invention
In order to solve the problems, the invention provides a visual detection method for the flatness of the surface of a basketball floor.
The visual detection method for the flatness of the basketball floor surface adopts the following technical scheme:
one embodiment of the invention provides a visual inspection method for the flatness of the surface of a basketball floor, which comprises the following steps:
collecting basketball floor images;
acquiring a Gaussian mixture model according to a gray level histogram of the basketball floor image, and acquiring a gray level value range of a normal area according to a mean value parameter and a standard deviation parameter of the Gaussian model in the Gaussian mixture model; taking the pixel points corresponding to all gray values outside the gray value range of the normal region in the gray histogram in the basketball floor image as initial defect pixel points of the basketball floor image; obtaining defective pixel points of the basketball floor image according to the initial defective pixel points of the basketball floor image;
taking each defective pixel point in the basketball floor image as each node, and marking any node in the basketball floor image and any node in eight adjacent nodes as a pair of adjacent nodes to obtain each pair of adjacent nodes; obtaining the offset degree between each adjacent node pair according to the distance between the gradient direction intersection point of two nodes in each adjacent node pair and the two nodes; acquiring a weighted undirected graph according to the offset degree between each adjacent node pair, and acquiring each initial category of the basketball floor image according to the weighted undirected graph; obtaining the circle center of each fitting circle according to the defect pixel points in each initial category of the basketball floor image, and obtaining each circle center category according to the circle centers of all fitting circles; taking defective pixel points in the initial category corresponding to all fitting circles of each circle center category as each final category of basketball floor images;
acquiring each defect area according to each final category of the basketball floor image; acquiring a gray level change curve of each defect area according to the gray level gradual change characteristic of each defect area; acquiring a characteristic value of each defect area according to the gray level change curve of each defect area; judging the type of each defect area according to the characteristic value of each defect area.
Preferably, the step of obtaining the gaussian mixture model according to the gray level histogram of the basketball floor image and obtaining the gray level range of the normal region according to the mean value parameter and the standard deviation parameter of the gaussian model in the gaussian mixture model comprises the following specific steps:
carrying out Gaussian mixture model fitting on a gray histogram of the basketball floor image by using an EM algorithm, wherein the number of sub-Gaussian models contained in the fitted Gaussian mixture model is 3, obtaining the mean value parameter and standard deviation parameter of each sub-Gaussian model, recording the sub-Gaussian model corresponding to the second large mean value parameter as a second sub-Gaussian model,is the normal region gray value range, wherein +.>Mean parameter representing the second sub-Gaussian model, < ->Representing standard deviation parameters of the second sub-gaussian model.
Preferably, the step of obtaining the defective pixel point of the basketball floor image according to the initial defective pixel point of the basketball floor image includes the following specific steps:
when the eight adjacent areas of any initial defective pixel point of the basketball floor image are provided with the initial defective pixel points, the initial defective pixel points are marked as defective pixel points, and all the defective pixel points of the basketball floor image are obtained.
Preferably, the step of obtaining the offset degree between each adjacent node pair includes the following specific steps:
marking any adjacent node pair as a current adjacent node pair; the intersection point of the gradient directions of two nodes in the current adjacent node pair is marked as a first intersection point;
in the method, in the process of the invention,representing a distance between a first node in the current adjacent node pair and a first intersection point; />Representing a distance between a second node in the current neighboring node pair and the first intersection point; />Representing the degree of offset between the current adjacent node pair; the degree of offset between all adjacent node pairs is obtained.
Preferably, the weighted undirected graph is obtained according to the offset degree between each adjacent node pair, and each initial category of the basketball floor image is obtained according to the weighted undirected graph, which comprises the following specific steps:
connecting each adjacent node pair in the basketball floor image to form an edge, and constructing a plurality of weighted undirected graphs by taking the offset degree between each adjacent node pair as an edge weight;
and clustering each weighted undirected graph by using a spectral clustering algorithm, dividing each weighted undirected graph into a plurality of subgraphs, and taking all corresponding nodes in each subgraph as each initial category of the basketball floor image.
Preferably, the center of each fitting circle is obtained according to the defective pixel point in each initial category of the basketball floor image, and each center category is obtained according to the centers of all fitting circles, comprising the following specific steps:
fitting circles to the defect pixel points in each initial category of the basketball floor image by using a least square method, and obtaining the circle center of each fitted circle; and clustering the circle centers of all the fitting circles by using a mean shift clustering algorithm to obtain each circle center category.
Preferably, the step of obtaining each defect area according to each final category of the basketball floor image includes the following specific steps:
and (3) performing convex hull detection on defective pixel points in each final category of the basketball floor image, and marking each obtained convex hull area as each defective area.
Preferably, the step of obtaining the gray scale variation curve of each defective area according to the gray scale gradation characteristic of each defective area includes the following specific steps:
the number of preset segmentsMarking any defect area as a current defect area, acquiring the mass center of the current defect area, marking the connection between the mass center of the current defect area and any edge point of the current defect area as a connection line, and marking the connection line as followsEach pixel point is divided into each segment, the gray average value of all the pixel points on each segment is obtained,numbering each section of the connecting line from the centroid of the current defect area, taking the number as an abscissa, taking the gray average value of each section of the connecting line as an ordinate, and drawing a gray change curve of the current defect area; and acquiring a gray level change curve of each defective area.
Preferably, the obtaining the characteristic value of each defect area according to the gray level change curve of each defect area includes the following specific steps:
in the method, in the process of the invention,a characteristic value representing a current defective area; />The gray level change curve representing the current defective area +.>Slope values of the data points; />Representing the number of data points in the gray level change curve of the current defect area; and acquiring the characteristic values of all the defect areas.
Preferably, the judging the type of each defect area according to the characteristic value of each defect area includes the following specific steps:
when the characteristic value of the defect area is positive, the defect area is a concave area, and when the characteristic value of the defect area is negative, the defect area is a convex area.
The technical scheme of the invention has the beneficial effects that: firstly, acquiring a Gaussian mixture model set according to a gray level histogram of a basketball floor image, and acquiring a gray level value range of a normal area according to a mean value parameter and a standard deviation parameter of a Gaussian model in the Gaussian mixture model so as to obtain an initial defect pixel point of the basketball floor image; then, each adjacent node pair is obtained according to the defective pixel point of the basketball floor image; obtaining the offset degree between each adjacent node pair according to the distance between the gradient direction intersection point of two nodes in each adjacent node pair and the two nodes, further obtaining a weighted undirected graph, and obtaining each initial category of the basketball floor image according to the weighted undirected graph; according to each initial category of the basketball floor image, the circle centers of each fitting circle are obtained, and the circle centers of all fitting circles are clustered to obtain each final category of the basketball floor image, so that defective pixel points in each final category of the basketball floor image belong to the same defective area, defective pixel points belonging to the concave area and the convex area are prevented from being clustered into one category, and the subsequent judgment of the category of the defective area is prevented from being influenced; finally, according to each final category of the basketball floor image, each defect area is obtained; acquiring a characteristic value of each defect area according to the gray gradient characteristic of each defect area, and judging that each defect area is a concave curve or a convex area according to the characteristic value of each defect area; the defect areas obtained by the invention are more complete, and the types of the defect areas are identified according to the gray level characteristics of the convex and concave defects, so that the staff can conveniently use different processing methods according to the types of the defect areas.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of steps of a visual inspection method for the flatness of a basketball floor surface according to the present invention;
FIG. 2 is a schematic illustration of a basketball floor gray scale image of the present invention;
FIG. 3 is a schematic representation of a gray level histogram of a basketball floor image of the present invention;
fig. 4 is a schematic diagram of a gray scale variation curve according to the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following detailed description refers to specific implementation, structure, characteristics and effects of a visual inspection method for the flatness of the basketball floor surface according to the invention, which is provided by the invention, with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following specifically describes a specific scheme of the visual inspection method for the flatness of the basketball floor surface provided by the invention with reference to the accompanying drawings.
Referring to fig. 1, a flowchart of steps of a visual inspection method for flatness of a basketball floor surface according to an embodiment of the invention is shown, the method includes the following steps:
s001, acquiring basketball floor images.
The basketball court used for the game is selected, the basketball court is clean and free of other sundries, an industrial camera is moved on the basketball court, basketball floor color images of different angles and positions are captured, the whole basketball court is covered as much as possible, strong light interference is avoided during shooting, in order to facilitate subsequent processing, the basketball floor color images are subjected to gray processing, and the images after gray processing are recorded as basketball floor images.
Thus, a basketball floor image is obtained, as shown in fig. 2, and fig. 2 is a schematic diagram of a basketball floor gray scale image according to the present invention.
S002, acquiring a Gaussian mixture model according to the gray level histogram of the basketball floor image, and acquiring a gray level range of a normal area according to the mean value parameter and the standard deviation parameter of the Gaussian mixture model in the Gaussian mixture model, so as to obtain the defective pixel point of the basketball floor image.
FIG. 3 is a schematic view of gray level histogram of a basketball floor image according to the present invention, as shown in FIG. 3.
It should be noted that, the unevenness of the basketball floor surface is caused by the surface bulge or the depression, the gray value of the depression area gradually increases from the depression center to the periphery under the influence of illumination, the gray value of the projection area gradually decreases from the center to the periphery, so that the gray value of the center area of the projection area is larger than the gray value of the normal area, the gray value of the peripheral area of the projection area is smaller than the gray value of the normal area, the gray value of the center area of the depression area is smaller than the gray value of the normal area, the gray value of the peripheral area of the depression area is larger than the gray value of the normal area, and the gray value distribution of the normal area is more consistent, so that three gray value characteristics are shared in the basketball floor image, the number of sub-gaussian models is obtained according to the three gray value characteristics of the basketball floor image, the gray histogram of the basketball floor image is fitted with the gaussian mixture model according to the number of the sub-gaussian models, and then the initial defect pixel points are obtained according to the gray distribution of each sub-gaussian model in the gaussian mixture model.
In the embodiment of the invention, a Gaussian mixture model fitting is carried out on a gray histogram of a basketball floor image by utilizing an Expectation-Maximization (EM) algorithm, which is abbreviated as EM algorithm, each sub-Gaussian model corresponds to one gray value characteristic, and the basketball floor image has three gray value characteristics, so that the number of the sub-Gaussian models contained in the fitted Gaussian mixture model is 3, and the mean value parameter and the standard deviation parameter of each sub-Gaussian model are obtained.
It should be noted that, the middle sub-gaussian model in the gaussian mixture model, that is, the sub-gaussian model corresponding to the second large-mean parameter is the sub-gaussian model corresponding to the pixel point of the normal region, and because in the gaussian distribution, most gray values are inIn the range, the basketball floor map is obtained according to the mean value and standard deviation of the second sub-Gaussian model in the Gaussian mixture modelAn initial defective pixel of the image.
In the embodiment of the invention, the sub-Gaussian model corresponding to the second large-mean parameter is recorded as a second sub-Gaussian model,is the gray value range of the non-initial defective pixel point, wherein +.>Mean parameter representing the second sub-Gaussian model, < ->And representing standard deviation parameters of the second sub-Gaussian model, and taking the pixel points corresponding to all gray values outside the gray value range of the normal region in the gray histogram in the basketball floor image as initial defect pixel points of the basketball floor image.
It should be noted that, after the initial defective pixel point of the basketball floor image is obtained, the initial defective pixel point may be an abnormal point caused by noise influence, or may be a defective pixel point in a concave or convex area, so that the influence of the noise point needs to be eliminated first.
In the embodiment of the invention, if the eight neighborhood of any initial defective pixel point of the basketball floor image exists, the initial defective pixel point is marked as a defective pixel point, otherwise, the initial defective pixel point is marked as an abnormal point, and all defective pixel points of the basketball floor image are obtained.
The method comprises the steps of obtaining a Gaussian mixture model according to a gray level histogram of a basketball floor image, obtaining a gray level range of a normal area according to a mean value parameter and a standard deviation parameter of the Gaussian mixture model in the Gaussian mixture model, and further obtaining defective pixel points of the basketball floor image.
S003, acquiring each adjacent node pair according to the defective pixel point of the basketball floor image, acquiring the offset degree between each adjacent node pair, and constructing a weighted undirected graph according to the offset degree between each adjacent node pair; carrying out spectral clustering on the weighted undirected graph to obtain each initial category of basketball floor images; obtaining the circle center of each fitting circle according to the defect pixel points in each initial category of the basketball floor image; and acquiring each final category of the basketball floor image according to the circle center of each fitting circle.
It should be noted that, the defect pixels of the basketball floor image obtained in the step S002 are known to be the pixels of the concave area or the pixels of the convex area, and the main purpose of the present invention is to identify the concave area and the convex area in the basketball floor image, so that the obtained defect pixels of the basketball floor image need to be classified into a plurality of categories, the defect pixels belonging to the concave area are classified into one category, the defect pixels belonging to the convex area are classified into one category, so that the defect categories in the basketball floor image are conveniently analyzed later to be the concave area or the convex area, and different types of defect areas are treated differently, but if the distance between the concave area and the convex area is close, the defect pixels of the concave area and the convex area may be clustered into one category after the defect pixels of the basketball floor image are clustered by using a clustering algorithm, so as to influence the judgment of the defect categories later.
It should be further noted that, under the influence of illumination, the gray value of the concave area gradually increases from the center of the concave to the periphery, and the gray value of the convex area gradually decreases from the center to the periphery, so that the gradient directions of the defect pixels belonging to the concave area are consistent, and the gradient directions of the defect pixels belonging to the convex area are consistent, and the gradient directions of the adjacent defect pixels are intersected.
In the embodiment of the invention, each defective pixel point in the basketball floor image is taken as each node, any node in the basketball floor image and any node in eight adjacent nodes are marked as adjacent node pairs, and all adjacent node pairs are obtained.
Marking any adjacent node pair as a current adjacent node pair; the intersection point of the gradient directions of two nodes in the current adjacent node pair is marked as a first intersection point; obtaining the offset degree between the current adjacent node pairs according to the first intersection point:
in the method, in the process of the invention,representing a distance between a first node in the current adjacent node pair and a first intersection point; />Representing a distance between a second node in the current neighboring node pair and the first intersection point; />Representing the degree of offset between the current adjacent node pair. The offset degree between all adjacent node pairs is obtained in the same way; />Representing absolute value symbols.
And taking each defective pixel point in the basketball floor image as each node, connecting each adjacent node pair in the basketball floor image to form an edge, and taking the offset degree between each adjacent node pair as an edge weight value to construct a plurality of weighted undirected graphs. The method comprises the steps of clustering each obtained weighted undirected graph by using a spectral clustering algorithm, dividing each weighted undirected graph into a plurality of subgraphs, wherein each subgraph comprises a plurality of nodes, and taking all the corresponding nodes in each subgraph as each initial category of basketball floor images.
It should be noted that, the defect pixel points in each initial category of the basketball floor image are similar to circles, so that the concave area or the convex area is formed by a plurality of concentric circles, and therefore, the circle centers of each fitting circle are obtained by fitting the defect pixel points in each initial category of the basketball floor image, and the circle centers of all fitting circles are clustered, so that the defect pixel points in each initial category corresponding to each fitting circle with the same circle center are pixel points of the same defect area, and the defect pixel points belonging to the concave area and the defect pixel points belonging to the convex area are prevented from being clustered into one category.
In the embodiment of the invention, a least square method is used for fitting circles to defective pixel points in each initial category of basketball floor images, and the circle center of each fitting circle is obtained; clustering the circle centers of all fitting circles by using a mean shift clustering algorithm to obtain each circle center category, and taking defective pixel points in an initial category corresponding to all fitting circles of each circle center category as each final category of the basketball floor image, wherein a least square method and the mean shift clustering algorithm are known techniques, and in the embodiment of the invention, excessive redundant description is not carried out.
To this end, each final category of basketball floor image is obtained.
S004, according to each final category of the basketball floor image, each defect area is obtained, a gray level change curve of each defect area is obtained, according to the gray level change curve of each defect area, a characteristic value of each defect area is obtained, and according to the characteristic value of each defect area, the defect type of each defect area is obtained, as shown in fig. 4, and fig. 4 is a schematic diagram of a gray level change curve of the invention.
It should be noted that, in the step S004, the defect pixel point in each final category in the basketball floor image is obtained, so in the embodiment of the present invention, in order to analyze the defect category later, each defect area needs to be obtained according to the defect pixel point in each final category.
In the embodiment of the invention, the convex hull detection is carried out on the defective pixel points in each final category of the basketball floor image, and each obtained convex hull area is recorded as each defective area.
It should be noted that, the gray value of the concave area gradually increases from the center of the concave to the periphery, and the gray value of the convex area gradually decreases from the center to the periphery, so that the convex or concave area is identified according to the gray value gradual change feature of each defect area, if the gray value from the center of mass of any defect area to the edge point of any defect area is gradually increased, the defect area is the concave area, and if the gray value from the center of mass of any defect area to the edge point of any defect area is gradually decreased, the defect area is the convex area.
In the embodiment of the invention, any defect area is marked as a current defect area, the mass center of the current defect area is obtained, the mass center of the current defect area is connected with any edge point of the current defect area and is marked as a connecting line, and the connecting line is used for connecting the current defect area according to the following steps ofDividing each pixel point into each segment, acquiring the gray average value of all the pixel points on each segment, numbering each segment of the connecting line from the center of mass of the current defect area, taking the number as an abscissa, taking the gray average value of each segment of the connecting line as an ordinate, and drawing a gray change curve of the current defect area, wherein the gray change curve is used for describing gray gradual change characteristics; in the embodiment of the invention, the number of segments is set +.>In other embodiments, the practitioner can set +.>Is a value of (2).
According to the gray level transformation curve of the current defect area, acquiring the characteristic value of the current defect area:
in the method, in the process of the invention,a characteristic value representing a current defective area; />The gray level change curve representing the current defective area +.>Slope values of the data points; />The number of data points in the gray level change curve representing the current defect area, when +.>And when the value of the defect area is larger, indicating that the current defect area is a concave area, and similarly, acquiring the characteristic value of each defect area.
When the characteristic value of any defect area is positive, the defect area is indicated to be a concave area, when the characteristic value of any defect area is negative, the defect area is indicated to be a convex area, and staff can use different repairing methods to process according to the type of the defect area, and when the characteristic value of any defect area is 0, the defect area is not indicated to be a concave area or a convex area.
So far, according to the defect pixel points in each final category of the basketball floor image, each defect area is obtained, the gray level change curve of each defect area is obtained, the characteristic value of each defect area is obtained according to the gray level change curve of each defect area, and the defect type of each defect area is obtained according to the characteristic value of each defect area.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the invention, but any modifications, equivalent substitutions, improvements, etc. within the principles of the present invention should be included in the scope of the present invention.
Claims (10)
1. The visual detection method for the flatness of the surface of the basketball floor is characterized by comprising the following steps of:
collecting basketball floor images;
acquiring a Gaussian mixture model according to a gray level histogram of the basketball floor image, and acquiring a gray level value range of a normal area according to a mean value parameter and a standard deviation parameter of the Gaussian model in the Gaussian mixture model; taking the pixel points corresponding to all gray values outside the gray value range of the normal region in the gray histogram in the basketball floor image as initial defect pixel points of the basketball floor image; obtaining defective pixel points of the basketball floor image according to the initial defective pixel points of the basketball floor image;
taking each defective pixel point in the basketball floor image as each node, and marking any node in the basketball floor image and any node in eight adjacent nodes as a pair of adjacent nodes to obtain each pair of adjacent nodes; obtaining the offset degree between each adjacent node pair according to the distance between the gradient direction intersection point of two nodes in each adjacent node pair and the two nodes; acquiring a weighted undirected graph according to the offset degree between each adjacent node pair, and acquiring each initial category of the basketball floor image according to the weighted undirected graph; obtaining the circle center of each fitting circle according to the defect pixel points in each initial category of the basketball floor image, and obtaining each circle center category according to the circle centers of all fitting circles; taking defective pixel points in the initial category corresponding to all fitting circles of each circle center category as each final category of basketball floor images;
acquiring each defect area according to each final category of the basketball floor image; acquiring a gray level change curve of each defect area according to the gray level gradual change characteristic of each defect area; acquiring a characteristic value of each defect area according to the gray level change curve of each defect area; judging the type of each defect area according to the characteristic value of each defect area.
2. The visual inspection method for the flatness of the surface of a basketball floor according to claim 1, wherein the steps of obtaining a gaussian mixture model according to a gray level histogram of an image of the basketball floor and obtaining a gray level range of a normal area according to a mean value parameter and a standard deviation parameter of a gaussian model in the gaussian mixture model are as follows:
carrying out Gaussian mixture model fitting on a gray histogram of the basketball floor image by using an EM algorithm, wherein the number of sub-Gaussian models contained in the fitted Gaussian mixture model is 3, obtaining the mean value parameter and standard deviation parameter of each sub-Gaussian model, recording the sub-Gaussian model corresponding to the second large mean value parameter as a second sub-Gaussian model,is the normal region gray value range, wherein +.>Mean parameter representing the second sub-Gaussian model, < ->Representing standard deviation parameters of the second sub-gaussian model.
3. The visual inspection method for the flatness of the surface of a basketball floor according to claim 1, wherein the step of obtaining defective pixels of the basketball floor image based on the initial defective pixels of the basketball floor image comprises the following specific steps:
when the eight adjacent areas of any initial defective pixel point of the basketball floor image are provided with the initial defective pixel points, the initial defective pixel points are marked as defective pixel points, and all the defective pixel points of the basketball floor image are obtained.
4. The visual inspection method for the flatness of a basketball floor surface according to claim 1, wherein the step of obtaining the degree of offset between each adjacent node pair comprises the steps of:
marking any adjacent node pair as a current adjacent node pair; the intersection point of the gradient directions of two nodes in the current adjacent node pair is marked as a first intersection point;
in the method, in the process of the invention,representing a distance between a first node in the current adjacent node pair and a first intersection point; />Representing a distance between a second node in the current neighboring node pair and the first intersection point; />Representing the degree of offset between the current adjacent node pair; the degree of offset between all adjacent node pairs is obtained.
5. The visual inspection method for the flatness of the surface of a basketball floor according to claim 1, wherein the steps of obtaining a weighted undirected graph according to the degree of deviation between each adjacent pair of nodes, and obtaining each initial category of the basketball floor image according to the weighted undirected graph, comprises the following specific steps:
connecting each adjacent node pair in the basketball floor image to form an edge, and constructing a plurality of weighted undirected graphs by taking the offset degree between each adjacent node pair as an edge weight;
and clustering each weighted undirected graph by using a spectral clustering algorithm, dividing each weighted undirected graph into a plurality of subgraphs, and taking all corresponding nodes in each subgraph as each initial category of the basketball floor image.
6. The visual inspection method for the flatness of basketball floor surface according to claim 1, wherein the steps of obtaining the center of each fitting circle according to the defective pixel point in each initial category of the basketball floor image, and obtaining each center category according to the centers of all fitting circles, comprise the following specific steps:
fitting circles to the defect pixel points in each initial category of the basketball floor image by using a least square method, and obtaining the circle center of each fitted circle; and clustering the circle centers of all the fitting circles by using a mean shift clustering algorithm to obtain each circle center category.
7. The visual inspection method for the flatness of the surface of a basketball floor according to claim 1, wherein the steps of obtaining each defective area according to each final category of the basketball floor image comprises the following steps:
and (3) performing convex hull detection on defective pixel points in each final category of the basketball floor image, and marking each obtained convex hull area as each defective area.
8. The visual inspection method for the flatness of basketball floor surface according to claim 1, wherein the step of obtaining the gray level change curve of each defective area according to the gray level gradual change characteristic of each defective area comprises the following specific steps:
the number of preset segmentsMarking any defect area as a current defect area, acquiring the mass center of the current defect area, marking the connection between the mass center of the current defect area and any edge point of the current defect area as a connecting line, and marking the connecting line as a +_ line>Dividing each pixel point into each segment, acquiring the gray average value of all the pixel points on each segment, numbering each segment of the connecting line from the centroid of the current defect area, taking the number as the abscissa, and connecting the connecting lineThe gray average value of each section of the current defect area is a vertical coordinate, and a gray change curve of the current defect area is drawn; and acquiring a gray level change curve of each defective area.
9. The visual inspection method for the flatness of a basketball floor surface according to claim 8, wherein the obtaining the characteristic value of each defective area according to the gray level change curve of each defective area comprises the following specific steps:
in the method, in the process of the invention,a characteristic value representing a current defective area; />The gray level change curve representing the current defective area +.>Slope values of the data points; />Representing the number of data points in the gray level change curve of the current defect area; and acquiring the characteristic values of all the defect areas.
10. The visual inspection method for the flatness of a basketball floor surface according to claim 1, wherein the judging of the type of each defective area based on the characteristic value of each defective area comprises the following specific steps:
when the characteristic value of the defect area is positive, the defect area is a concave area, and when the characteristic value of the defect area is negative, the defect area is a convex area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311596693.0A CN117314902B (en) | 2023-11-28 | 2023-11-28 | Visual inspection method for flatness of basketball floor surface |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311596693.0A CN117314902B (en) | 2023-11-28 | 2023-11-28 | Visual inspection method for flatness of basketball floor surface |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117314902A true CN117314902A (en) | 2023-12-29 |
| CN117314902B CN117314902B (en) | 2024-01-30 |
Family
ID=89288691
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311596693.0A Active CN117314902B (en) | 2023-11-28 | 2023-11-28 | Visual inspection method for flatness of basketball floor surface |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117314902B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106780440A (en) * | 2016-11-29 | 2017-05-31 | 北京邮电大学 | Destruction circuit plate relic image automatic comparison recognition methods |
| CN109523541A (en) * | 2018-11-23 | 2019-03-26 | 五邑大学 | A kind of metal surface fine defects detection method of view-based access control model |
| US20220198647A1 (en) * | 2021-02-09 | 2022-06-23 | Nanjing University Of Aeronautics And Astronautics | Method for detecting and recognizing surface defects of automated fiber placement composite based on image converted from point cloud |
-
2023
- 2023-11-28 CN CN202311596693.0A patent/CN117314902B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106780440A (en) * | 2016-11-29 | 2017-05-31 | 北京邮电大学 | Destruction circuit plate relic image automatic comparison recognition methods |
| CN109523541A (en) * | 2018-11-23 | 2019-03-26 | 五邑大学 | A kind of metal surface fine defects detection method of view-based access control model |
| US20220198647A1 (en) * | 2021-02-09 | 2022-06-23 | Nanjing University Of Aeronautics And Astronautics | Method for detecting and recognizing surface defects of automated fiber placement composite based on image converted from point cloud |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117314902B (en) | 2024-01-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113538433B (en) | Mechanical casting defect detection method and system based on artificial intelligence | |
| US20220148213A1 (en) | Method for fully automatically detecting chessboard corner points | |
| CN113658132B (en) | Computer vision-based structural part weld joint detection method | |
| CN102525381B (en) | The recording equipment of image processing apparatus, image processing method and embodied on computer readable | |
| CN114170228B (en) | Computer image edge detection method | |
| CN115082467A (en) | Building material welding surface defect detection method based on computer vision | |
| CN103543394A (en) | Discharge ultraviolet imaging quantization parameter extraction method of high-voltage electric equipment | |
| CN107248161A (en) | Retinal vessel extracting method is supervised in a kind of having for multiple features fusion | |
| CN115345885A (en) | Method for detecting appearance quality of metal fitness equipment | |
| CN109493339B (en) | Endoscope imaging-based method for detecting defects of pores on inner surface of casting | |
| CN115311277B (en) | Pit defect identification method for stainless steel product | |
| CN105675625A (en) | Method for detecting fruit surface defects by virtue of segmentation of gradient iteration threshold | |
| CN111553873B (en) | Automatic detection method for brain neurons based on multi-scale convolution neural network | |
| CN107230207A (en) | The detection method and system of tire | |
| CN114494256A (en) | Electric wire production defect detection method based on image processing | |
| CN115797361B (en) | Aluminum template surface defect detection method | |
| CN115049667B (en) | Gear defect detection method | |
| CN115880280B (en) | Method for detecting quality of welding seam of steel structure | |
| CN115082477B (en) | Semiconductor wafer processing quality detection method based on light reflection removing effect | |
| CN116883408B (en) | Integrating instrument shell defect detection method based on artificial intelligence | |
| CN115908774B (en) | Quality detection method and device for deformed materials based on machine vision | |
| CN113313107A (en) | Intelligent detection and identification method for multiple types of diseases on cable surface of cable-stayed bridge | |
| CN115170570A (en) | Fabric fuzzing and pilling detection method based on gray level run-length matrix | |
| CN117314902B (en) | Visual inspection method for flatness of basketball floor surface | |
| CN112750113B (en) | Glass bottle defect detection method and device based on deep learning and linear detection |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |