CN115049667A - Gear defect detection method - Google Patents
Gear defect detection method Download PDFInfo
- Publication number
- CN115049667A CN115049667A CN202210978401.9A CN202210978401A CN115049667A CN 115049667 A CN115049667 A CN 115049667A CN 202210978401 A CN202210978401 A CN 202210978401A CN 115049667 A CN115049667 A CN 115049667A
- Authority
- CN
- China
- Prior art keywords
- point
- gear
- target
- edge
- defect
- 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
- 230000007547 defect Effects 0.000 title claims abstract description 203
- 238000001514 detection method Methods 0.000 title claims abstract description 46
- 238000000034 method Methods 0.000 claims description 23
- 238000004364 calculation method Methods 0.000 claims description 8
- 239000000126 substance Substances 0.000 claims description 6
- 238000007781 pre-processing Methods 0.000 claims description 5
- 238000003708 edge detection Methods 0.000 claims description 4
- 238000005299 abrasion Methods 0.000 abstract description 2
- 238000005259 measurement Methods 0.000 description 4
- 208000002697 Tooth Abrasion Diseases 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 238000001914 filtration Methods 0.000 description 3
- 208000004188 Tooth Wear Diseases 0.000 description 2
- 206010044048 Tooth missing Diseases 0.000 description 2
- 230000002950 deficient Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000008451 emotion Effects 0.000 description 1
- 238000011326 mechanical measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 101150049349 setA gene Proteins 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
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
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20024—Filtering details
- G06T2207/20032—Median filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20112—Image segmentation details
- G06T2207/20164—Salient point detection; Corner detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30164—Workpiece; Machine component
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Quality & Reliability (AREA)
- Length Measuring Devices By Optical Means (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of gear defect detection, in particular to a gear defect detection method, which comprises the following steps: acquiring a surface image of a gear to be detected, and further acquiring a gear area image of the surface image; according to the gear area image, determining the gear edge, the gear center point and each real angular point, further determining each suspected angular point and each pseudo angular point, further determining a distance defect index value, an area defect index value and an angle defect index value of the gear, combining the three defect index values, calculating a comprehensive defect index value of the gear, and finally determining the defect grade of the gear. The invention can accurately determine the defect degree of the gear by detecting the defects of gear abrasion, uneven pitch, missing teeth, broken teeth and the like, and effectively improves the accuracy of gear defect detection.
Description
Technical Field
The invention relates to the technical field of gear defect detection, in particular to a gear defect detection method.
Background
As a common and important transmission part, gears are widely used in various mechanical devices. The quality of the gear is crucial to the whole equipment, and once the gear adopted by the mechanical equipment has a large defect, the gear is possibly damaged under the conditions of heavy load and excessive stress, even the whole mechanical equipment is possibly failed and scrapped, and the life safety of operators is threatened when the gear is serious. Therefore, it is particularly important to periodically detect the defects of the gears during the working process.
The traditional gear mechanical measurement and detection method belongs to a contact detection method, is inconvenient to operate, is easy to damage the surface of a gear, is low in measurement efficiency, is easy to cause measurement errors or errors due to factors such as emotion and fatigue of detection personnel, and is poor in detection accuracy.
Disclosure of Invention
The invention aims to provide a gear defect detection method which is used for solving the problem of poor accuracy of the existing gear defect detection.
In order to solve the technical problem, the invention provides a gear defect detection method, which comprises the following steps:
acquiring a surface image of a gear to be detected, and preprocessing the surface image to obtain a gear area image of the surface image;
determining the edge of the gear, the center point of the gear and each real angular point according to the gear area image;
determining each suspected corner point according to the gear edge and each real corner point, and determining each pseudo corner point according to the gear edge and the gear center point;
determining a distance value from each corner point to the central point according to the central point, each real corner point, each suspected corner point and each pseudo corner point of the gear, and further determining a distance defect index value of the gear according to the distance value from each corner point to the central point;
determining each first edge target point and each second edge target point according to the gear edge, each real angular point and each pseudo angular point, and further determining each target triangle;
determining the filling degree of each target triangle according to each target triangle and the gear area image, and determining the area defect index value of the gear according to the filling degree of each target triangle;
determining each target angle value according to each first edge target point, each second edge target point and the gear center point, and determining an angle defect index value of the gear according to each target angle value;
and calculating a comprehensive defect index value of the gear according to the distance defect index value, the area defect index value and the angle defect index value of the gear, and determining the defect grade of the gear according to the comprehensive defect index value of the gear.
Further, the determining each suspected corner point includes:
carrying out Hough circle detection on each real angular point so as to obtain Hough circles of inner angular points;
and determining each intersection point of the gear edge and the Hough circle of the inner side corner points, and removing the intersection points belonging to the real corner points from each intersection point so as to obtain each suspected corner point.
Further, the determining each pseudo corner point includes:
for any edge pixel point a2 on the edge of the gear, determining a connecting line between the any edge pixel point a2 and the center point of the gear and a vertical line segment passing through the edge pixel point a 2;
determining a directed line segment formed by any edge pixel point a2 and the adjacent edge pixel point a1 on one side of the any edge pixel point a2
The included angle between the vertical line segment and the
And a directed line segment formed by any edge pixel point a2 and the adjacent edge pixel point a3 on the other side of the edge pixel point a2
The included angle between the vertical line segment and the
；
If the included angle is
And an included angle
And if the edge pixel point a2 has a different sign, determining that the any edge pixel point a2 is a pseudo corner point.
Further, the determining a distance defect index value of the gear comprises:
classifying the distance values from each corner point to the center point according to the distance values from each corner point to the center point so as to obtain each distance category, and determining each normal category and each defect category in each distance category according to the number of the distance values from each corner point to the center point in each distance category;
carrying out Hough circle detection on each real angular point so as to obtain Hough circles of the outer angular points;
calculating the ratio of the distance value from each corner point to the center point in each defect category to the radius of the Hough circle of the outer corner points, and accumulating all the ratios corresponding to each defect category to obtain a sub-distance defect index value corresponding to each defect category;
calculating the variance of the distance values from all the corners to the central point in each normal category according to the distance values from the corners to the central point in each normal category, thereby obtaining the sub-distance defect index values corresponding to each normal category;
and calculating the distance defect index value of the gear according to the sub distance defect index value corresponding to each defect type and the sub distance defect index value corresponding to each normal type.
Further, a calculation formula corresponding to the distance defect index value of the gear is calculated as follows:
wherein the content of the first and second substances,
is the distance defect index value of the gear,
is the distance defect weight value corresponding to the defect class,
is the distance defect weight value corresponding to the normal class,
and
sub-distance defect index values corresponding to the ith defect class and the jth normal class respectively,
and
the total number of defect classes and normal classes, respectively.
Further, the determining each first edge target point and each second edge target point, and further determining each target triangle, includes:
carrying out Hough circle detection on each real angular point so as to obtain an outer angular point Hough circle, and carrying out edge detection on each pseudo angular point so as to obtain a pseudo angular point Hough circle;
determining the intersection point of the edge of the gear and the Hough circle of the outer angular point to obtain each first intersection point, and determining the intersection point of the edge of the gear and the Hough circle of the pseudo angular point to obtain each second intersection point;
performing density clustering on each first intersection point to obtain each first intersection point category, and calculating the position coordinate mean value of all the first intersection points in each first intersection point category to obtain each first edge target point;
performing density clustering on each second intersection point to obtain each second intersection point category, and calculating the position coordinate mean value of all second intersection points in each second intersection point category to obtain each second edge target point;
and determining target first edge target points corresponding to any two adjacent second edge target points according to the positions of the second edge target points and the first edge target points, and connecting the any two adjacent second edge target points and the corresponding target first edge target points pairwise to obtain each target triangle.
Further, the determining the target first edge target points corresponding to any two adjacent second edge target points includes:
according to the second edge target points and the positions of the first edge target points, if a certain first edge target point is located on the inner side of any two adjacent second edge target points, the certain first edge target point is used as a target first edge target point corresponding to any two adjacent second edge target points, and if no certain first edge target point is located on the inner side of any two adjacent second edge target points, a first edge target point which is located on the outer side of any two adjacent second edge target points and is closest to the center point of any two adjacent second edge target points is used as a target first edge target point corresponding to any two adjacent second edge target points.
Further, the determining the filling degree of each target triangle and determining the area defect index value of the gear according to the filling degree of each target triangle includes:
determining the area of each target triangle and the area occupied by the gear in each target triangle according to the image of each target triangle and the gear area;
calculating the ratio of the area occupied by the gear in each target triangle to the area of the target triangle, thereby obtaining the filling degree of each target triangle;
and calculating the variance of the filling degree of each target triangle so as to obtain the index value of the area defect of the gear.
Further, the determining each target angle value and determining an angle defect index value of the gear according to each target angle value includes:
connecting each first edge target point with the gear center point according to each first edge target point, each second edge target point and the gear center point, simultaneously connecting each second edge target point with the gear center point, and obtaining each target angle value according to an included angle formed by any two adjacent connecting lines;
and determining an angle fluctuation approximate entropy according to each target angle value, and calculating an angle defect index value of the gear according to the angle fluctuation approximate entropy.
Further, a calculation formula corresponding to the calculation of the angle defect index value of the gear is as follows:
wherein the content of the first and second substances,
is an index value of the angle defect of the gear,
the entropy is approximated for angular fluctuations.
The invention has the following beneficial effects: the method comprises the steps of obtaining a surface image of a gear to be detected, extracting gear angular points based on the surface image to obtain real angular points, suspected angular points and pseudo angular points on the gear, analyzing the angular points to determine distance defect index values, area defect index values and angle defect index values of the gear, measuring defects such as gear tooth abrasion, tooth pitch unevenness, tooth missing, tooth breakage and the like, and finally accurately determining the defect degree of the gear without contact measurement, so that the accuracy of gear defect detection is effectively improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions and advantages of the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a flow chart of a gear defect detection method in an embodiment of the present invention;
FIG. 2 is a grayscale image of a gear region in an embodiment of the invention;
FIG. 3 is a schematic illustration of a gear tooth shape in an embodiment of the present invention;
FIG. 4 is a schematic diagram illustrating the presence of a first edge target point inside two adjacent second edge target points in an embodiment of the present invention;
FIG. 5 is a schematic diagram illustrating the absence of a first edge target point inside two adjacent second edge target points according to an embodiment of the present invention.
Detailed Description
To further explain the technical means and effects of the present invention adopted to achieve the predetermined objects, the following detailed description of the embodiments, structures, features and effects of the technical solutions according to the present invention will be given with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" refers to not necessarily the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
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 embodiment provides a gear defect detection method, which is based on a surface image of a gear, can detect defects such as gear tooth abrasion, uneven tooth pitch, missing tooth, broken tooth and the like, and compared with the traditional measurement, the method can realize non-contact detection, is convenient and efficient, and has accurate results. Specifically, a flow chart corresponding to the gear defect detection method is shown in fig. 1, and includes the following steps:
(1) the method comprises the steps of obtaining a surface image of a gear to be detected, and preprocessing the surface image to obtain a gear area image of the surface image.
When the gear needs to be subjected to defect detection, the gear to be detected is shot through the CCD camera, so that a surface image of the gear to be detected is obtained, and the surface image is a color image in an RGB format. The obtained surface image is preprocessed, and the preprocessing process comprises the steps of carrying out gear region extraction, graying, equalization, boundary enhancement, filtering and the like on the surface image, so that the influence of a background region is eliminated, the noise of the image is removed, the quality of the image is improved, the image becomes clearer, the identifiability of the image is enhanced, the image feature extraction is facilitated, and the accuracy of a final defect detection result is improved.
In this embodiment, a weighted average method is used to perform graying processing on the image without the background region, so as to obtain a more reasonable gear grayscale image, and fig. 2 shows the grayscale image of the gear region in this embodiment. In order to remove the image noise and retain the boundary information of the image, the present embodiment selects the median filter to perform the filtering process on the image. Since the specific implementation process of performing the processes of gear region extraction, graying, equalization, boundary enhancement, filtering, and the like on the surface image belongs to the prior art, the detailed description is omitted here. Through this preprocessing process, a gear area image of the face image can be obtained.
(2) And determining the gear edge, the gear center point and each real angular point according to the gear area image.
As shown in fig. 2, the gear to be detected in the present embodiment includes a plurality of gear teeth, and as shown in fig. 3, each gear tooth is formed by combining an isosceles trapezoid 1 and a rectangle 2, and an arc line is formed between the gear teeth. And (3) performing edge detection on the gear area image on the basis of the step (1) to obtain the gear edge of the gear area image.
Meanwhile, Harris angular point detection is adopted to detect angular points in the gear area image, and the detected angular points are called as real angular points. The detected angular points actually comprise the angular points corresponding to the gear teeth and the angular points in the middle of the gear, and the subsequent gear defect detection only needs to consider the angular points corresponding to the gear teeth, so that the detected angular points are screened, and only the angular points corresponding to the gear teeth are reserved, so that the final real angular points are obtained. Under the normal condition without abrasion, four angular points can be extracted from each gear tooth, and the four angular points refer to four vertexes of an isosceles trapezoid respectively.
In addition, the gear area image is subjected to center pixel point detection, so that a gear center point can be obtained. In this embodiment, in order to obtain the center point of the gear, a voting threshold G1 is set according to the final position information of each real corner point, and hough circle fitting is performed, so that two hough circles can be obtained, where one hough circle is a hough circle fitted to the outer real corner point and one hough circle fitted to the inner real corner point, and the two hough circles are referred to as an outer corner point hough circle and an inner corner point hough circle, respectively, and the radius of the outer corner point hough circle is greater than the radius of the inner corner point hough circle. And calculating the coordinate mean value of the center coordinates of the hough circle of the outer side angular point and the hough circle of the inner side angular point, and taking the point corresponding to the coordinate mean value as the central point of the gear.
(3) And determining each suspected corner point according to the gear edge and each real corner point, and determining each pseudo corner point according to the gear edge and the gear center point.
The location where the corner point was originally present may become smooth due to gear tooth wear and other possible factors, and may not be detected by the corner point detection. For this problem, in this embodiment, on the basis of the step (2), each suspected corner point is determined according to the gear edge and each real corner point, and the specific implementation process includes:
and (3.1) carrying out Hough circle detection on each real angular point, thereby obtaining the Hough circle of the inner angular point.
Referring to the step (2), hough circle detection is performed on each real angular point to obtain two hough circles, and the hough circle with the smaller radius in the two hough circles is used as the hough circle of the inner angular point.
And (3.2) determining each intersection point of the gear edge and the Hough circle of the inner side corner point, and removing the intersection points belonging to the real corner points from each intersection point so as to obtain each suspected corner point.
And (3) determining each intersection point of the gear edge and the inner corner point Hough circle on the basis of the step (3.1), extracting newly formed intersection points of the gear edge and the inner corner point Hough circle from the intersection points, and calling the newly formed intersection points as suspected corner points. That is, intersection points belonging to real corner points are removed from these intersection points, and the remaining intersection points are regarded as suspected corner points. These suspected corner points may be due to gear tooth wear and possibly other factors, where the corner points were originally smooth and could not be detected by the corner point detection.
After each suspected corner point is determined, in order to facilitate the subsequent detection of the gear defect, each pseudo corner point on the gear is further determined according to the gear edge and the gear center point, and the specific implementation process comprises the following steps:
(3.3) for any edge pixel point a2 on the edge of the gear, determining a connecting line of the any edge pixel point a2 and the center point of the gear and a vertical line segment passing through the edge pixel point a 2.
(3.4) determining a directed line segment formed by any edge pixel point a2 and the adjacent edge pixel point a1 on one side of the any edge pixel point a2
The included angle between the vertical line segment and the
And a directed line segment formed by any edge pixel point a2 and the adjacent edge pixel point a3 on the other side of the edge pixel point a2
The included angle between the vertical line segment and the
；
(3.5) if said angle is included
And an included angle
And if the edge pixel point a2 has a different sign, determining that the any edge pixel point a2 is a pseudo corner point.
For facilitating understanding of the above steps (3.3) - (3.5), for any three adjacent edge pixel points a1 (x 1, y 1), a2 (x 2, y 2) and a3 (x 3, y 3) on the edge of the gear, connecting an edge pixel point a2 (x 2, y 2) with the center point of the gear, making a perpendicular line of the connection line through the edge pixel point a2 (x 2, y 2), and simultaneously connecting a1 (x 1, y 1) and a2 (x 2, y 2) to obtain a directional line segment
Connecting a2 (x 2, y 2) and a3 (x 3, y 3) to obtain a directed line segment
Is provided with a directional line segment
At an angle to the vertical of
Directed line segment
At an angle to the vertical of
If the clockwise included angle is positive, the counterclockwise included angle is negative
And
one is positive and one is negative, i.e. one is positive
And
opposite sign, in which there is a line segment
And
and located on both sides of the vertical line, the edge pixel point a2 (x 2, y 2) is defined as a pseudo corner point. In this way, all edge pixel points on the gear edge are traversed, so that all edge pixel points on the gear edge can be foundA pseudo corner point. The pseudo corner point is actually a certain pixel point on a connecting arc line between any two gear teeth, and the tangent of the pixel point at the arc line is perpendicular to the connecting line of the pixel point and the central point of the gear.
(4) And determining a distance value from each corner point to the central point according to the gear central point, each real corner point, each suspected corner point and each pseudo corner point, and further determining a distance defect index value of the gear according to the distance value from each corner point to the central point.
After the gear center point, each real angular point, each suspected angular point and each pseudo angular point are determined through the steps (2) and (3), the distance values from each real angular point, each suspected angular point and each pseudo angular point to the gear center point are determined, the distance values are called angular point-to-center point distance values, and the distance defect index value of the gear is determined based on the distance values, wherein the specific implementation process comprises the following steps:
and (4.1) classifying the distance values from each corner point to the central point according to the distance values from each corner point to the central point so as to obtain each distance category, and determining each normal category and each defect category in each distance category according to the number of the distance values from each corner point to the central point in each distance category.
Clustering operation is performed on the distance value from each corner point to the central point by using a density clustering algorithm, in this embodiment, clustering operation is performed on the distance value from each corner point to the central point by using a DBSCAN algorithm, so that each distance category is obtained. Before clustering operation is carried out on the distance value from each corner point to the central point by adopting a DBSCAN algorithm, a distance threshold value eps and a density threshold value MinPts are firstly set, and the setting of two parameters of the distance threshold value eps and the density threshold value MinPts depends on the size of a specific part. Assuming that the gear region image specification is 500 × 500 pixels, the distance threshold eps =15 and the density threshold MinPts =2 are set.
After clustering operation is carried out through the DBSCAN algorithm, each distance class can be obtained. Because the number of normal corners is usually large, and the abnormal corners are only extremely individual corners, the threshold value G for the number of corners is set, and at this time, for any distance class e, the distance class e is setThe number G of the distance values from each corner point to the center point in the distance class e is compared with the threshold value G of the number of corner points, if
Then the distance class e is taken as the normal class if
Then the distance class e is taken as the defect class. By the method, each distance category can be screened, so that each normal category and each defect category are obtained, the number of the normal categories is recorded as n, and the number of the defect categories is recorded as m.
And (4.2) carrying out Hough circle detection on each real angular point, thereby obtaining the Hough circle of the outer angular point.
Referring to the step (2), hough circle detection is performed on each real angular point, so that two hough circles can be obtained, and the hough circle with the larger radius in the two hough circles is used as the hough circle of the outer angular point.
And (4.3) calculating the ratio of the distance value from each corner point to the center point in each defect type to the radius of the Hough circle of the outer corner points, and accumulating all the ratios corresponding to each defect type to obtain the sub-distance defect index value corresponding to each defect type.
And (4) calculating the ratio of the distance value from each corner point to the center point in each defect type to the radius of the Hough circle of the outer corner points for the m defect types obtained in the step (4.1), summing the ratios, and taking the value obtained after summation as a sub-distance defect index value corresponding to the defect type.
And (4.4) calculating the variance of the distance values from all the corners to the central point in each normal category according to the distance values from the corners to the central point in each normal category, thereby obtaining the sub-distance defect index values corresponding to each normal category.
And (4) calculating the mean value of the distance values from all the corner points to the central point in each normal category for the n normal categories obtained in the step (4.1), calculating the variance of the distance values from all the corner points to the central point based on the mean value, and taking the calculated variance as the index value of the sub-distance defect corresponding to the normal category.
(4.5) calculating the distance defect index value of the gear according to the sub-distance defect index value corresponding to each defect type and the sub-distance defect index value corresponding to each normal type, wherein the corresponding calculation formula is as follows:
wherein the content of the first and second substances,
is the distance defect index value of the gear,
is the distance defect weight value corresponding to the defect class,
for the distance defect weight value corresponding to the normal category, the distance defect weight value corresponding to the defect category is set in this embodiment
=0.1, distance defect weight value corresponding to normal class
=0.9，
And
sub-distance defect index values corresponding to the ith defect class and the jth normal class respectively,
and
the total number of defect classes and normal classes, respectively.
(5) Determining each first edge target point and each second edge target point according to the gear edge, each real angular point and each pseudo angular point, and further determining each target triangle, wherein the specific implementation steps comprise:
and (5.1) carrying out Hough circle detection on each real angular point so as to obtain an outer angular point Hough circle, and carrying out edge detection on each pseudo angular point so as to obtain a pseudo angular point Hough circle.
Referring to the step (2), hough circle detection is performed on each real angular point to obtain two hough circles, and the hough circle with the larger radius in the two hough circles is used as the hough circle of the outer angular point. Meanwhile, Hough circle detection is carried out on each pseudo corner point, so that the Hough circle of the pseudo corner point is obtained.
And (5.2) determining the intersection points of the gear edge and the Hough circle of the outer corner point to obtain each first intersection point, and determining the intersection points of the gear edge and the Hough circle of the pseudo corner point to obtain each second intersection point.
And extracting intersection points of the Hough circle fitted to the corner points at the outermost end of the gear, namely the Hough circle of the corner points at the outer side and the edge of the gear, and calling the extracted intersection points as first intersection points. Meanwhile, intersection points of the hough circle of the pseudo corner points and the edge of the gear are extracted, and the extracted intersection points are called second intersection points.
(5.3) carrying out density clustering on each first intersection point to obtain each first intersection point category, and calculating the position coordinate mean value of all the first intersection points in each first intersection point category to obtain each first edge target point.
And (5) performing density clustering on each first intersection according to the spatial position on the basis of the step (5.2), so as to obtain each first intersection type. Wherein the intersection in each first intersection category refers to a real corner point on one gear tooth, and since the number of real corner points available on each gear tooth is 2 in case of no wear of the gear tooth, and the number of real corner points available on each gear tooth may be 0 or 1 in case of wear of the gear tooth, the number of intersection in each first intersection category is 1 or 2.
After each first intersection point category is obtained, for each first intersection point category, calculating a mean value of position coordinates of all first intersection points in the first intersection point category, and taking a point corresponding to the mean value as a first edge target point. According to the position information, recording the first edge target point set corresponding to all the first intersection point types as
。
(5.4) carrying out density clustering on the second intersection points to obtain second intersection point categories, and calculating the position coordinate mean value of all the second intersection points in the second intersection point categories to obtain second edge target points.
And (5) performing density clustering on each second intersection point according to the spatial position on the basis of the step (5.2), so as to obtain each second intersection point type. Wherein, the nodical point in every second nodical point classification means the marginal pixel point on the pitch arc between two adjacent teeth of a cogwheel, because under the condition that the pitch arc does not take place wearing and tearing, can obtain a marginal pixel point on every pitch arc, and this marginal pixel point is the pseudo-angular point on the pitch arc, and under the condition that the inside caving in of wearing and tearing takes place in the pitch arc, the marginal pixel point number that every pitch arc can be obtained just can be 2, therefore the number of the intersection point in every second nodical point classification is 1 or 2.
After each second intersection point category is obtained, for each second intersection point category, calculating the mean value of the position coordinates of all second intersection points in the second intersection point category, and taking the point corresponding to the mean value as a second edge target point. According to the position information, recording the second edge target point sets corresponding to all the second intersection point types as
。
And (5.5) determining target first edge target points corresponding to any two adjacent second edge target points according to the positions of the second edge target points and the first edge target points, and connecting the any two adjacent second edge target points and the corresponding target first edge target points pairwise to obtain each target triangle.
Considering that the number of the second edge target points is equal to the number of the arcs between two adjacent gear teeth, that is, the arcs between any two adjacent gear teeth correspond to one second edge target point, and because of the existence of the wear condition, the number of the first edge target points may be smaller than the number of the gear teeth, in order to obtain each target triangle, it is first necessary to determine the target first edge target points corresponding to any two adjacent second edge target points, and then perform pairwise connection on any two adjacent second edge target points and the corresponding target first edge target points, so as to obtain each target triangle. The specific process of determining the target first edge target point corresponding to any two adjacent second edge target points is as follows:
according to the positions of the second edge target points and the first edge target points, if a certain first edge target point is located on the inner side of any two adjacent second edge target points, the certain first edge target point is used as a target first edge target point corresponding to any two adjacent second edge target points, and if no certain first edge target point is located on the inner side of any two adjacent second edge target points, the first edge target point which is located on the outer side of any two adjacent second edge target points and is closest to the center point of any two adjacent second edge target points is used as a target first edge target point corresponding to any two adjacent second edge target points.
As shown in FIG. 4, for any two adjacent second edge target points
And
if there is a first edge target point
At the second edge target point
And
i.e. from the first edge target point
To a second edge target point
And
the formed line segment is taken as a vertical line, and the formed vertical foot is positioned at a second edge target point
And
on the constructed line segment, we then consider the first edge-target point
At a second edge target point
And
at this time, the first edge target point is set
As two adjacent second edge target points
And
target first edge target point.
As shown in FIG. 5, for any two adjacent second edge target points
And
if there is no first edge target point located at the second edge target point
And
is caused by the second edge target point
And
the central gear tooth is worn to cause that the corresponding angular point can not be extracted from the gear tooth, and then the central gear tooth is positioned at the second edge target point
And
is outside and is distant from the second edge target point
And
first edge target point with the nearest center point
As a second edge target point
And
target first edge target point.
And after the target first edge target points corresponding to any two adjacent second edge target points are obtained, constructing a triangle by taking the connecting line of any two adjacent second edge target points as a bottom edge and the corresponding target first edge target point as a vertex, wherein the triangle is the target triangle.
(6) And determining the filling degree of each target triangle according to each target triangle and the gear area image, and determining the area defect index value of the gear according to the filling degree of each target triangle.
On the basis of the step (5), determining the area of each target triangle and the area occupied by the gear in each target triangle according to each target triangle and the gear area image; calculating the ratio of the area occupied by the gear in each target triangle to the area of the target triangle, thereby obtaining the filling degree of each target triangle; and calculating the variance of the filling degree of each target triangle so as to obtain the index value of the area defect of the gear.
That is, for each target triangle, the area of the target triangle is determined and is denoted as s, the area occupied by the inner teeth of the target triangle is determined and is denoted as t, and then the filling degree u = t/s of the target triangle is obtained. And calculating the variance of the filling degree of all the target triangles, and taking the variance as the area defect index value H of the gear. When the number of the collapse angles of the gear is larger, the area defect index value H of the corresponding gear is larger, and the gear defect is more serious.
(7) Determining each target angle value according to each first edge target point, each second edge target point and the gear center point, and determining an angle defect index value of the gear according to each target angle value, wherein the specific implementation steps comprise:
(7.1) connecting each first edge target point with the gear center point according to each first edge target point, each second edge target point and the gear center point, simultaneously connecting each second edge target point with the gear center point, and obtaining each target angle value according to an included angle formed by any two adjacent connecting lines.
And respectively connecting each first edge target point and each second edge target point with the central point of the gear, wherein any two adjacent connecting lines form an included angle, and the angle value of the included angle is the target angle value. By the method, each target angle value can be obtained, and a target angle value sequence can be obtained
。
And (7.2) determining an angle fluctuation approximate entropy according to each target angle value, and calculating an angle defect index value of the gear according to the angle fluctuation approximate entropy.
On the basis of the above step (7.1), if the gear is not defective, then all the target angle values are the same, and the sequence of target angle values
No fluctuation is generated. If the angular point of the gear tooth is deviated or even lacks the gear tooth, the target angle value sequence
Fluctuation is generated, the larger the angular point deviation degree is, the more teeth are missing, and then the target angle value sequence is
The greater the ripple produced, the higher the complexity and the more severe the gear defect. In this embodiment, to measure the sequence of target angle values
Calculating the sequence of target angle values
The approximate entropy of (2) may also be referred to as angle fluctuation approximate entropy, and since the specific process of calculating the sequence approximate entropy belongs to the prior art, it is not described herein again. Normalizing the calculated angle fluctuation approximate entropy to obtain an angle defect index value of the gear, wherein the corresponding calculation formula is as follows:
wherein the content of the first and second substances,
is an index value of the angle defect of the gear,
the entropy is approximated for angular fluctuations.
(8) And calculating a comprehensive defect index value of the gear according to the distance defect index value, the area defect index value and the angle defect index value of the gear, and determining the defect grade of the gear according to the comprehensive defect index value of the gear.
And finally, integrating the distance defect index value, the area defect index value and the angle defect index value of the gear, calculating the integrated defect index value of the gear to evaluate the defect degree of the gear, wherein the corresponding calculation formula is as follows:
wherein the content of the first and second substances,
is the comprehensive defect index value of the gear,
、
and
respectively a distance defect index value, an area defect index value and an angle defect index value of the gear,
、
and
the weights corresponding to the distance defect index value, the area defect index value, and the angle defect index value, respectively, in this embodiment,
，
，
。
and after obtaining the comprehensive defect index value of the gear, determining the defect grade of the gear according to the comprehensive defect index value of the gear. Specifically, if the comprehensive defect index value of the gear is smaller than the first defect index threshold value, judging that the gear is free of defects; if the comprehensive defect index value of the gear is not less than the first defect index threshold value and less than the second defect index threshold value, judging the gear to be a light defect; if the comprehensive defect index value of the gear is not less than the second defect index threshold value and less than the third defect index threshold value, judging that the gear is a common defect; and if the comprehensive defect index value of the gear is not less than the third defect index threshold value, judging that the gear is a serious defect.
Wherein the first defect indicator threshold, the second defect indicator threshold, and the third defect indicator threshold may be based on gear defectsThe specific tolerance level. In this embodiment, the first defect index threshold value is set to 0.1, the second defect index threshold value is set to 0.4, and the third defect index threshold value is set to 0.7. At this time, when
If so, judging that the gear is not defective; when in use
Judging the gear to be a light defect; when it is 0.
Judging that the gear is a common defect; when in use
And if so, judging the gear to be a serious defect. Thus, the defect detection of the gear is completed.
According to the invention, the gear angular points are extracted based on the surface image of the gear to be detected, so that each real angular point, suspected angular point and pseudo angular point on the gear are obtained, and the angular points are analyzed, so that the distance defect index value, the area defect index value and the angle defect index value of the gear can be determined, the three defect index values can be used for measuring the defects such as gear tooth abrasion, tooth pitch unevenness, tooth missing, tooth breakage and the like, finally, the defect degree of the gear can be accurately determined, and the accuracy of gear defect detection is effectively improved.
It should be noted that: the above-mentioned embodiments are only used for illustrating the technical solutions of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; such modifications and substitutions do not substantially depart from the spirit and scope of the embodiments of the present application and are intended to be included within the scope of the present application.
Claims (10)
1. A gear defect detection method is characterized by comprising the following steps:
acquiring a surface image of a gear to be detected, and preprocessing the surface image to obtain a gear area image of the surface image;
determining the edge of the gear, the center point of the gear and each real angular point according to the gear area image;
determining each suspected angular point according to the gear edge and each real angular point, and determining each pseudo angular point according to the gear edge and the gear center point;
determining a distance value from each corner point to the central point according to the central point, each real corner point, each suspected corner point and each pseudo corner point of the gear, and further determining a distance defect index value of the gear according to the distance value from each corner point to the central point;
determining each first edge target point and each second edge target point according to the gear edge, each real angular point and each pseudo angular point, and further determining each target triangle;
determining the filling degree of each target triangle according to each target triangle and the gear area image, and determining the area defect index value of the gear according to the filling degree of each target triangle;
determining each target angle value according to each first edge target point, each second edge target point and the gear center point, and determining an angle defect index value of the gear according to each target angle value;
and calculating a comprehensive defect index value of the gear according to the distance defect index value, the area defect index value and the angle defect index value of the gear, and determining the defect grade of the gear according to the comprehensive defect index value of the gear.
2. The method of claim 1, wherein the determining each suspected corner point comprises:
carrying out Hough circle detection on each real angular point so as to obtain Hough circles of inner angular points;
and determining each intersection point of the gear edge and the Hough circle of the inner side corner points, and removing the intersection points belonging to the real corner points from each intersection point so as to obtain each suspected corner point.
3. The gear defect detecting method of claim 1, wherein the determining each pseudo corner point comprises:
for any edge pixel point a2 on the edge of the gear, determining a connecting line between the any edge pixel point a2 and the center point of the gear and a vertical line segment passing through the edge pixel point a 2;
determining a directed line segment formed by any edge pixel point a2 and the adjacent edge pixel point a1 on one side of the any edge pixel point a2
The included angle between the vertical line segment and the
And a directed line segment formed by any edge pixel point a2 and the adjacent edge pixel point a3 on the other side of the edge pixel point a2
The included angle between the vertical line segment and the
；
If the included angle is
And an included angle
And if the edge pixel point a2 is different from the original edge pixel point, determining that the edge pixel point a2 is a pseudo corner point.
4. The gear defect detection method of claim 1, wherein the determining a distance defect index value for the gear comprises:
classifying the distance values from each corner point to the center point according to the distance values from each corner point to the center point so as to obtain each distance category, and determining each normal category and each defect category in each distance category according to the number of the distance values from each corner point to the center point in each distance category;
carrying out Hough circle detection on each real angular point so as to obtain Hough circles of the outer angular points;
calculating the ratio of the distance value from each corner point to the center point in each defect category to the radius of the Hough circle of the outer corner points, and accumulating all the ratios corresponding to each defect category to obtain a sub-distance defect index value corresponding to each defect category;
calculating the variance of the distance values from all the corners to the central point in each normal category according to the distance values from the corners to the central point in each normal category, thereby obtaining sub-distance defect index values corresponding to each normal category;
and calculating the distance defect index value of the gear according to the sub-distance defect index value corresponding to each defect type and the sub-distance defect index value corresponding to each normal type.
5. The gear defect detection method according to claim 4, wherein the calculation formula for calculating the distance defect index value of the gear is as follows:
wherein the content of the first and second substances,
is a distance defect index value of the gear,
is the distance defect weight value corresponding to the defect class,
is the distance defect weight value corresponding to the normal class,
and
sub-distance defect index values corresponding to the ith defect class and the jth normal class respectively,
and
the total number of defect classes and normal classes, respectively.
6. The gear defect detection method of claim 1, wherein the determining each first edge target point and each second edge target point, and further determining each target triangle, comprises:
carrying out Hough circle detection on each real angular point so as to obtain an outer angular point Hough circle, and carrying out edge detection on each pseudo angular point so as to obtain a pseudo angular point Hough circle;
determining the intersection point of the edge of the gear and the Hough circle of the outer angular point to obtain each first intersection point, and determining the intersection point of the edge of the gear and the Hough circle of the pseudo angular point to obtain each second intersection point;
performing density clustering on each first intersection point to obtain each first intersection point category, and calculating the position coordinate mean value of all the first intersection points in each first intersection point category to obtain each first edge target point;
performing density clustering on each second intersection point to obtain each second intersection point category, and calculating the position coordinate mean value of all second intersection points in each second intersection point category to obtain each second edge target point;
and determining target first edge target points corresponding to any two adjacent second edge target points according to the positions of the second edge target points and the first edge target points, and connecting the any two adjacent second edge target points and the corresponding target first edge target points pairwise to obtain each target triangle.
7. The gear defect detecting method according to claim 6, wherein the determining the target first edge target points corresponding to any two adjacent second edge target points comprises:
according to the second edge target points and the positions of the first edge target points, if a certain first edge target point is located on the inner side of any two adjacent second edge target points, the certain first edge target point is used as a target first edge target point corresponding to any two adjacent second edge target points, and if no certain first edge target point is located on the inner side of any two adjacent second edge target points, a first edge target point which is located on the outer side of any two adjacent second edge target points and is closest to the center point of any two adjacent second edge target points is used as a target first edge target point corresponding to any two adjacent second edge target points.
8. The gear defect detecting method according to claim 6, wherein the determining the filling degree of each target triangle and the determining the area defect index value of the gear according to the filling degree of each target triangle comprises:
determining the area of each target triangle and the area occupied by the gear in each target triangle according to the image of each target triangle and the gear area;
calculating the ratio of the area occupied by the gear in each target triangle to the area of the target triangle, thereby obtaining the filling degree of each target triangle;
and calculating the variance of the filling degree of each target triangle so as to obtain the index value of the area defect of the gear.
9. The method for detecting the gear defect according to claim 1, wherein the determining each target angle value and determining the angle defect index value of the gear according to each target angle value comprises:
connecting each first edge target point with the gear center point according to each first edge target point, each second edge target point and the gear center point, simultaneously connecting each second edge target point with the gear center point, and obtaining each target angle value according to an included angle formed by any two adjacent connecting lines;
and determining an angle fluctuation approximate entropy according to each target angle value, and calculating an angle defect index value of the gear according to the angle fluctuation approximate entropy.
10. The method for detecting the gear defect according to claim 8, wherein the calculation formula for calculating the angle defect index value of the gear is as follows:
wherein, the first and the second end of the pipe are connected with each other,
is an index value of the angle defect of the gear,
the entropy is approximated for angular fluctuations.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210978401.9A CN115049667B (en) | 2022-08-16 | 2022-08-16 | Gear defect detection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210978401.9A CN115049667B (en) | 2022-08-16 | 2022-08-16 | Gear defect detection method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115049667A true CN115049667A (en) | 2022-09-13 |
| CN115049667B CN115049667B (en) | 2022-12-06 |
Family
ID=83166431
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210978401.9A Active CN115049667B (en) | 2022-08-16 | 2022-08-16 | Gear defect detection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115049667B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115239723A (en) * | 2022-09-22 | 2022-10-25 | 江苏东晨机械科技有限公司 | Plate rolling machine part detection method based on pattern recognition |
| CN116862908A (en) * | 2023-09-01 | 2023-10-10 | 菏泽学院 | Metal gear meshing defect detection method based on image processing |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20190002946A (en) * | 2017-06-30 | 2019-01-09 | 조선대학교산학협력단 | Method for evaluating the reliability of hybrid large curvic gear by infrared thermography |
| CN111915572A (en) * | 2020-07-13 | 2020-11-10 | 青岛大学 | Self-adaptive gear pitting quantitative detection system and method based on deep learning |
| CN113658133A (en) * | 2021-08-16 | 2021-11-16 | 江苏鑫丰源机电有限公司 | Gear surface defect detection method and system based on image processing |
| CN114494045A (en) * | 2022-01-10 | 2022-05-13 | 南京工大数控科技有限公司 | Large-scale straight gear geometric parameter measuring system and method based on machine vision |
| CN114723701A (en) * | 2022-03-31 | 2022-07-08 | 南通博莹机械铸造有限公司 | Gear defect detection method and system based on computer vision |
-
2022
- 2022-08-16 CN CN202210978401.9A patent/CN115049667B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20190002946A (en) * | 2017-06-30 | 2019-01-09 | 조선대학교산학협력단 | Method for evaluating the reliability of hybrid large curvic gear by infrared thermography |
| CN111915572A (en) * | 2020-07-13 | 2020-11-10 | 青岛大学 | Self-adaptive gear pitting quantitative detection system and method based on deep learning |
| CN113658133A (en) * | 2021-08-16 | 2021-11-16 | 江苏鑫丰源机电有限公司 | Gear surface defect detection method and system based on image processing |
| CN114494045A (en) * | 2022-01-10 | 2022-05-13 | 南京工大数控科技有限公司 | Large-scale straight gear geometric parameter measuring system and method based on machine vision |
| CN114723701A (en) * | 2022-03-31 | 2022-07-08 | 南通博莹机械铸造有限公司 | Gear defect detection method and system based on computer vision |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115239723A (en) * | 2022-09-22 | 2022-10-25 | 江苏东晨机械科技有限公司 | Plate rolling machine part detection method based on pattern recognition |
| CN115239723B (en) * | 2022-09-22 | 2022-12-23 | 江苏东晨机械科技有限公司 | Plate rolling machine part detection method based on pattern recognition |
| CN116862908A (en) * | 2023-09-01 | 2023-10-10 | 菏泽学院 | Metal gear meshing defect detection method based on image processing |
| CN116862908B (en) * | 2023-09-01 | 2023-11-24 | 菏泽学院 | Metal gear meshing defect detection method based on image processing |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115049667B (en) | 2022-12-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115049667B (en) | Gear defect detection method | |
| CN114723701B (en) | Gear defect detection method and system based on computer vision | |
| CN115082418B (en) | Precise identification method for automobile parts | |
| CN107239742B (en) | Method for calculating scale value of instrument pointer | |
| WO2021168733A1 (en) | Defect detection method and apparatus for defect image, and computer-readable storage medium | |
| CN113379712B (en) | Steel bridge bolt disease detection method and system based on computer vision | |
| CN115375686B (en) | Glass edge flaw detection method based on image processing | |
| CN115690108A (en) | Aluminum alloy rod production quality evaluation method based on image processing | |
| CN115311267B (en) | Method for detecting abnormity of check fabric | |
| CN115861307B (en) | Fascia gun power supply driving plate welding fault detection method based on artificial intelligence | |
| CN116523923B (en) | Battery case defect identification method | |
| CN105139384B (en) | The method and apparatus of defect capsule detection | |
| CN116703890B (en) | Method and system for detecting tab defects | |
| US8542229B2 (en) | Identification method of data point distribution area on coordinate plane and recording medium | |
| CN115290663B (en) | Mini LED wafer appearance defect detection method based on optical detection | |
| JP5123244B2 (en) | Shape defect inspection device, shape modeling device, and shape defect inspection program | |
| CN112669295A (en) | Lithium battery pole piece defect detection method based on secondary threshold segmentation theory | |
| CN114972575A (en) | Linear fitting algorithm based on contour edge | |
| CN115619708A (en) | Method for judging fault based on image recognition of oil quality change of main shaft bearing lubricating oil | |
| EP2973223B1 (en) | Enhanced analysis for image-based serpentine belt wear evaluation | |
| CN115272350A (en) | Method for detecting production quality of computer PCB mainboard | |
| CN117152129B (en) | Visual detection method and system for surface defects of battery cover plate | |
| CN114170168A (en) | Display module defect detection method, system and computer readable storage medium | |
| JP5067677B2 (en) | Defect detection method, defect detection apparatus, and program | |
| CN111627053B (en) | Method and system for detecting cleaning cleanliness of universal spinneret orifice |
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 | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: A gear defect detection method Granted publication date: 20221206 Pledgee: Jiangsu Qidong rural commercial bank Co.,Ltd. cooperative sub branch Pledgor: Qidong Qunhe Machinery Equipment Co.,Ltd. Registration number: Y2024980012451 |