CN115047006A - Diode surface quality detection method - Google Patents
Diode surface quality detection method Download PDFInfo
- Publication number
- CN115047006A CN115047006A CN202210548200.5A CN202210548200A CN115047006A CN 115047006 A CN115047006 A CN 115047006A CN 202210548200 A CN202210548200 A CN 202210548200A CN 115047006 A CN115047006 A CN 115047006A
- Authority
- CN
- China
- Prior art keywords
- glass bulb
- area
- diode
- defect
- images
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/95—Investigating the presence of flaws or contamination characterised by the material or shape of the object to be examined
Landscapes
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The invention relates to the technical field of material testing and analysis, in particular to a diode surface quality detection method. The method obtains visible light images and corresponding mapping images of the transparent diode at multiple angles through an optical means and specifically utilizes the visible light means, carries out analysis and test on the surface of the transparent diode glass bulb according to the visible light images and the corresponding mapping images, can eliminate the judgment influence of the defects of the internal structural part of the diode on the real defects of the surface of the glass bulb and the judgment influence of the reflection of the glass bulb on the real defects of the surface of the glass bulb through the optical means when judging the surface quality of the transparent diode glass bulb, improves the identification accuracy of the real defects of the surface of the transparent diode glass bulb, and solves the problem that the surface quality of the transparent diode cannot be accurately detected because the surface defects of the transparent diode cannot be accurately identified.
Description
Technical Field
The invention relates to the technical field of material testing and analysis, in particular to a diode surface quality detection method.
Background
The light emitting diode is one of the most commonly used light emitting diodes, and the package casing is usually made of colorless or colored transparent glass material, which is called glass envelope. For the light emitting diode, because the glass bulb of the diode is transparent, and the diode has an internal structure and a profile characteristic, the interference of the profile characteristic of an internal structural component of the diode and the defect characteristic of the internal structural component can be easily caused in the process of identifying the surface defect of the glass bulb of the light emitting diode in an optical mode, and meanwhile, the glass bulb of the diode also has a light reflection phenomenon, the interference can be caused in the process of identifying the surface defect of the glass bulb of the light emitting diode in the optical mode, so that the detection of the surface defect of the transparent glass bulb of the light emitting diode is very difficult at present, and the surface quality detection of the transparent diode cannot be accurately finished.
Therefore, a method for effectively identifying the surface defects of the transparent diode glass bulb is needed, and the defects are accurately positioned, so that the surface defects of the diode are accurately identified.
Disclosure of Invention
The invention provides a diode surface quality detection method, which is used for solving the problem that the prior art cannot accurately identify the surface defects of a light-transmitting diode glass shell, and adopts the following technical scheme:
the invention discloses a diode surface quality detection method, which comprises the following steps:
vertically placing the transparent diode, horizontally irradiating the transparent diode by using a light source, rotating the transparent diode by using a set rotating angle as an interval, and shooting the complete circumference of the transparent diode by using a visible light means to obtain an even number of transparent diode glass bulb surface images and an even number of mapping images corresponding to the glass bulb surface images; the glass bulb surface image is shot at the same side of the light source, and the mapping image corresponding to the glass bulb surface image is shot at the opposite side of the light source;
performing linear detection on each glass bulb surface image, selecting any one pixel point on all determined straight lines as an initial point, determining a structural part area in the corresponding glass bulb surface image by a seed filling method, taking an area except the structural part area in the glass bulb surface image as a non-structural part area, and determining the structural part area and the non-structural part area in the corresponding mapping image according to a mapping relation; performing edge identification on the surface image of the glass envelope, removing structural part area outline edge information and glass envelope external outline edge information from the identified edge information, determining a defect area in the surface image of the glass envelope according to the residual edge information, and determining a defect area in a corresponding mapping image according to a mapping relation;
dividing two glass bulb surface images in the symmetrical direction and two corresponding mapping images into the same symmetrical group, wherein the shooting angles of the two glass bulb surface images in each symmetrical group are just opposite and the shooting angles of the two corresponding mapping images are just opposite, determining each overlapping area obtained by a defect area in the non-structural part area of the two mapping images in the symmetrical group, judging whether the ratio of the area of each overlapping area to the area of the two defect areas of the obtained overlapping area is larger than a set ratio, and if so, considering that the two defect areas of the obtained overlapping area are real defect areas; the two defect areas of the obtained overlapping area are defect areas corresponding to the positions of the overlapping area in the two mapping images; calculating the average gray value of the defect areas corresponding to the real defect area positions on the two glass bulb surface images in the symmetrical group, wherein the defect area with the smaller average gray value is the real defect in the non-structural part area on the glass bulb surface image;
in structural part areas of two glass bulb surface images of the symmetrical group, determining whether the gray gradient direction of each defect area is vertical to the axis of the transparent diode glass bulb, if not, determining that the corresponding defect area is a real defect in the structural part area on the glass bulb surface image;
obtaining the real defects on the surface image of each glass bulb according to the real defects in the non-structural part area and the real defects in the structural part area on the surface image of each glass bulb, matching and connecting the surface images of the glass bulbs under any two adjacent shooting angles, and determining the number and the size of the real defects on the complete circumference of the transparent diode glass bulb;
and finishing the surface quality evaluation of the transparent diode according to the number and the size of the real defects on the complete circumference of the glass bulb.
The invention has the beneficial effects that:
the invention obtains a plurality of glass bulb surface images and corresponding mapping images on the complete circumference of the transparent diode by using a visible light means, analyzes and tests the surface of the transparent diode glass bulb according to the obtained plurality of glass bulb surface images and the corresponding mapping images, can eliminate the judgment influence of the defects of the internal structural part of the diode on the real defects of the surface of the glass bulb by using an optical means when judging the surface quality of the transparent diode glass bulb, and eliminates the judgment influence of the reflection of the glass bulb on the real defects of the surface of the glass bulb, thereby improving the identification accuracy of the real defects of the surface of the transparent diode glass bulb and realizing the accurate detection of the surface quality of the transparent diode.
Further, determining the topological relation of each real defect in the graph structure of each glass bulb surface image, and completing the matching and connection of the glass bulb surface images at any two adjacent shooting angles according to the topological relation of each real defect in the graph structure of each glass bulb surface image.
Further, the specific method for completing the evaluation of the surface quality of the transparent diode by using the number and the size of the real defects on the complete circumference of the glass bulb comprises the following steps:
calculating the surface quality degradation value of the diode:
wherein the content of the first and second substances,
representing the number of real defects, S representing the total area of the real defects, and S representing the total area of the surface of the glass envelope; the larger the q value, the lower the transparent diode surface quality evaluation.
Further, the set ratio is 90%, the set rotation angle is 45 °, and the number of the even number of transparent diode glass envelope surface images is 8.
Drawings
FIG. 1 is a flow chart of the diode surface quality detection method of the present invention.
Fig. 2 is a schematic structural diagram of a light emitting diode according to the present invention.
Fig. 3 is a schematic view of the shooting angle of the present invention.
Fig. 4 is a schematic diagram of the light path of the light of the present invention from the light source to the inner structure of the led and then out of the structure.
Fig. 5 is a schematic diagram of gray values of a row of pixel points in a matrix corresponding to each defect region in a structural member region on a surface image of the glass envelope.
Fig. 6 is a schematic diagram of the present invention showing the surface image of the glass bulb photographed at two adjacent photographing angles.
Fig. 7 is a schematic representation of the nearest four nodes in the graph structure of the envelope surface image of the present invention.
Fig. 8 is a schematic diagram of four nodes adjacent to node a when node a is targeted according to the present invention.
Fig. 9 is a schematic diagram of the corresponding matching relationship of the same real defect in the graph structures of different glass envelope surface images in the graph structures of the glass envelope surface images obtained at two adjacent shooting angles according to the present invention.
Fig. 10 is a schematic view of the position of the same real defect of the present invention in the surface images of the envelope obtained at a plurality of consecutive photographing angles.
Detailed Description
The following describes a method for detecting the surface quality of a diode according to the present invention in detail with reference to the accompanying drawings and embodiments.
The method comprises the following steps:
the overall flow of the embodiment of the diode surface quality detection method is shown in fig. 1, and the specific process is as follows:
1. and collecting the surface images of the diode glass shells in different directions and corresponding mapping images.
The led is the most common transparent diode, so the present embodiment takes the led as an example for analysis. As shown in fig. 2, it is a schematic structural diagram of a light emitting diode, and a structural member for emitting light after being energized is packaged in a transparent glass envelope, and is connected to an external circuit by positive and negative pin pins with different lengths.
In this embodiment, in order to accurately probe the surface defects of the light emitting diode glass bulb, the surface image of the diode glass bulb and the corresponding mapping image are collected in a darkroom environment. Specifically, the diode is vertically placed, the diode is horizontally irradiated by the light source, the surface image of the glass bulb of the diode is shot by the first camera on the same side of the light source, a piece of white paper is placed on the opposite side of the light source, the diode forms a mapping image corresponding to the surface image of the glass bulb on the white paper under the action of the light source, and the mapping image on the white paper is shot and recorded by the second camera on the opposite side of the light source.
In order to accurately identify the defect of a complete circle of the diode glass bulb, the surface images of the glass bulb and the corresponding mapping images at different angles are shot again after the diode is rotated, and finally 2N glass bulb surface images and the corresponding mapping images on the circle of the diode glass bulb are obtained in a mode of continuously rotating the diode and changing the angle of the diode. It is easy to understand that, the smaller the rotation angle of the diode is in two adjacent shots, the more the total number of 2N glass bulb surface images and corresponding mapping images obtained after the diode rotates for one circle, and in this embodiment, in order to consider the defect identification efficiency and the accuracy, the rotation angle value is set to 45 ° each time, so as to form 2N =8 shooting directions as shown in fig. 3, that is, a total of eight glass bulb surface images and corresponding mapping images are obtained.
2. And identifying a defect area on the surface image of the glass bulb, and specifically subdividing the collected surface image of the glass bulb into a structural part area and a non-structural part area.
As shown in fig. 2, based on the structural feature that the light emitting diode is connected to the internal structural component by two needle pins, in this embodiment, after graying the surface image of the glass envelope to obtain the surface grayscale image of the glass envelope, hough line detection is performed on the surface grayscale image of the glass envelope to determine two straight lines, and then any one pixel point on the two straight lines is used as an initial point to determine the structural component region of the glass envelope portion, that is, the structural component region of the surface image of the glass envelope, by a seed filling method.
Then, the determined structural part area is removed from the glass bulb area in the glass bulb surface image, and the non-structural part area of the glass bulb surface image can be correspondingly obtained.
Because the mapping relation exists between the glass bulb surface image and the corresponding mapping image, after the structural part area and the non-structural part area of the glass bulb surface image are determined, the structural part area and the non-structural part area in the mapping image can be correspondingly obtained.
Meanwhile, CANNY edge detection is carried out on the surface gray level image of the glass bulb, and because texture information does not exist on the surface of the glass bulb of a normal light-emitting diode when the surface of the glass bulb has no defects, if texture information still exists after the removal of the outline of the structural part, namely the boundary of the structural part area, and the outer outline of the glass bulb is removed, the texture information can be preliminarily considered as defect information, and a plurality of defect areas can be correspondingly obtained.
In practice, however, the defect information does not necessarily represent defects on the surface of the glass envelope, and the defect information includes three possibilities: a. true defects on the surface of the glass envelope; b. a false defect formed by the change of light shadow caused by the reflection action of the glass bulb; c. defects in the internal structural components of the envelope.
Since accurate identification of defects on the surface of the envelope is to be achieved, it is necessary to identify and eliminate the defects in cases b and c.
3. And calculating the defect response of each defect region in the non-structural part region on the mapping image corresponding to the surface image of the glass envelope, and determining the real defects in the non-structural part region on the surface image of the glass envelope.
In step 1, 2N glass envelope surface images and corresponding mapping images can be obtained in total by rotating the diode at different angles, and then if two glass envelope surface images and corresponding mapping images which are mutually in symmetrical directions are drawn into the same symmetrical group, N symmetrical groups can be formed in total, wherein each symmetrical group comprises two symmetrical glass envelope surface images with just opposite shooting angles, namely symmetrical glass envelope surface images, and corresponding and same symmetrical mapping images of the two symmetrical glass envelope surface images.
In the present embodiment, since the shooting angle is 2N =8, N =4 symmetrical groups are obtained, specifically referring to fig. 3, i.e., 0 ° -180 °, 45 ° -225 °, 90 ° -270 °, 135 ° -315 °.
In each symmetrical group of two symmetrical mapping images which are symmetrical to each other, removing pixel points of the structural part area and pixel points outside the defect area, then judging the overlapping part of the defect area in the two symmetrical mapping images to obtain a plurality of overlapping areas, and if the ratio of the area of a certain overlapping area to the area of the defect area of the obtained overlapping area is greater than a set ratio, considering that the defect area of the obtained overlapping area is a real defect area.
The defect area of the overlapping area is obtained by referring to the defect area corresponding to the position of the overlapping area in the two mapping images, and the selected set ratio is 90%.
The real defect regions determined above are actually two defect regions corresponding to positions in two mapping images that are symmetric to each other on a symmetric set, but actually one defect often exists only on one side, and the defect region is recognized on the other mapping image only because the envelope is transparent. Therefore, it is also necessary to determine, in the two symmetrical glass bulb surface images of the symmetrical group, which region has a smaller average gray value in the two regions corresponding to the positions of the real defect regions, and the region having the smaller average gray value is the region where the real defect is located.
From this, the specific distribution position, number and size of all real defects in the non-structural member region on all 2N =8 envelope surface images can be determined.
4. And calculating the illumination distribution characteristics of each defect area in the structural part area on the surface image of the glass bulb, and determining the real defects in the structural part area on the surface image of the glass bulb.
As shown in fig. 4, the diode is a structure with a ring glass in a middle vacuum when viewed from above, and the ring glass, i.e. the glass envelope, has a shielding effect although the light transmittance is large, and particularly for a colored light emitting diode, the light blocking effect of the colored glass is larger.
In fig. 4, the horizontal line in the inner circle represents the internal structure of the glass envelope, the light is reflected back after reaching the position, and the straight lines o, a, and b represent different light paths, respectively.
On the light paths o, a and b, the thicknesses of the glass envelopes through which the light rays need to pass are sequentially increased, so that the brightness of the light rays reaching the structural member through the three light paths and reflected out is sequentially reduced based on the blocking effect of the glass envelopes on the light rays.
The rule that the brightness of light rays which are emitted into the glass envelope and then emitted out of the glass envelope gradually decreases from the central position of the glass envelope to two sides is along the horizontal direction, so that the horizontal direction needs to be found at first, and the direction vertical to a straight line obtained through Hough line detection is taken as the horizontal direction. Moreover, even if the glass bulb has a light reflection phenomenon, the law that the brightness of light rays which are injected into the glass bulb and then are injected out of the glass bulb is gradually reduced from the central position of the glass bulb to two sides is still established under the condition that the light reflection phenomenon exists because the whole brightness of the area is improved by the light reflection phenomenon; similarly, the rule that the brightness of the light rays emitted into the glass envelope is gradually reduced from the center of the glass envelope to the two sides is still true even if the structural member has defects because the structural member has defects, the reflection effect of the structural member on the light rays cannot be influenced, and the reflection of the light rays cannot be blocked.
It can be determined that the rule that the brightness of light rays emitted into the glass envelope is gradually reduced from the central position of the glass envelope to the two sides is broken only when the thickness change rule of the partial area of the glass envelope does not conform to the gradual increase from the central position of the glass envelope to the two sides because the surface of the glass envelope has defects.
Therefore, whether the defect area is generated by the real defect on the surface of the glass bulb can be judged by judging whether the illumination distribution characteristics of each defect area in the structural part area on the surface image of the glass bulb conform to the rule.
The illumination distribution characteristics of each defect area in the structural part area on the glass bulb surface image are determined by judging the gray level change condition of each defect area in the horizontal direction, and specifically, the illumination distribution characteristics of the defect areas are analyzed by adopting a principal component analysis method as follows:
each defect area has a certain range, and can be regarded as a matrix formed by finite rows and finite columns of pixel points, and the pixel point value change condition on each row of the matrix is calculated.
(1) And establishing a coordinate system, wherein the interval of the abscissa is 1, and the ordinate is the gray value of each pixel point.
If the gray values of a certain row of pixel points in the matrix are sequentially as follows: [1, 2, 1, 2], the coordinate representation obtained is shown in FIG. 5, with the abscissa used for counting and the ordinate used for gray scale values.
(2) And calculating the main direction of the gray value of each row of pixel points in the defect area by a principal component analysis method.
And obtaining principal component directions of the data by using a principal component analysis algorithm according to the discrete point coordinates of the gray values of the pixel points on each row in the matrix, wherein if the matrix has K rows, K principal component directions can be correspondingly obtained, and each principal component direction is a 2-dimensional unit vector.
(3) And determining the standard main direction of each defect area in the structural part area on the surface image of the glass envelope when the real defect does not exist on the surface of the glass envelope.
When the surface of the glass shell has no real defects, the gray value of the pixel points in each line in each defect area in the structural part area on the surface image of the glass shell has a single variation trend, because the brightness of the light at the central position of the glass shell, namely the light path o shown in the attached drawing 4, is the maximum, when the defect area is on the left side of the central position of the glass shell, the gray value of the pixel points in each line in the defect area is sequentially increased along the horizontal direction, correspondingly, in the coordinates shown in the attached drawing 5, each discrete point sequentially rises along the positive direction of the horizontal coordinate, the standard main direction of the gray value of the pixel points in the corresponding line is on the upper side along the positive direction of the horizontal coordinate, and the specific upper angle of the gray value can be determined in advance according to the detected light transmittance of the glass shell of the light-emitting diode.
When the defect area is on the right side of the center position of the glass bulb, the gray values of the pixel points in each row in the defect area are sequentially reduced along the horizontal direction, correspondingly, in the coordinate shown in the attached drawing 5, each discrete point is sequentially reduced along the positive direction of the horizontal coordinate, the standard main direction of the gray values of the pixel points in the corresponding row is deviated along the positive direction of the horizontal coordinate, and the specific downward angle can be determined in advance according to the detected light transmittance of the glass bulb of the light-emitting diode.
(4) And judging the deviation degree of the main direction of the gray value of each row of pixel points in the defect area and the standard main direction of the defect area.
And comparing the K main component directions obtained in each defect area in the structural part area on the glass bulb surface image with the standard main direction of the defect area, and judging whether the main component directions with set proportion in the K main component directions are the same as the standard main direction, if so, determining that the defect area in the structural part area on the glass bulb surface image does not have the real defect of the glass bulb surface, and if not, determining that the defect area in the structural part area on the glass bulb surface image has the real defect of the glass bulb surface.
In other embodiments, the gray gradient direction of each defect area in the structural component area on the glass bulb surface image can be directly calculated, and then whether the gray gradient direction is horizontal or not is judged, if so, the defect area in the structural component area on the glass bulb surface image is considered to have no real defect of the glass bulb surface, and if not, the defect area in the structural component area on the glass bulb surface image is considered to have real defect of the glass bulb surface.
Therefore, the specific distribution positions, the number and the sizes of all real defects in the structural part area on all the 2N =8 glass bulb surface images can be determined, and the specific distribution positions, the number and the sizes of all the real defects in all the glass bulb surface images can be determined by combining the specific distribution positions, the number and the sizes of all the real defects in the non-structural part area on all the 2N =8 glass bulb surface images determined in the step 3.
5. And calculating the matching relation of the real defects in the surface images of the glass envelope at any two adjacent angles, and determining the correct number and size of the real defects on the complete circle of the surface of the glass envelope.
Through the steps 1-4, the real defects in the surface images of the glass bulb at each shooting angle can be judged, but actually, the real defects in the surface images of the glass bulb at two adjacent shooting angles have a plurality of repetitions, and the number of the real defects in the surface images of the glass bulb at all angles is far higher than the actual number of the real defects.
This is because after two adjacent shooting angles are rotated by 45 °, images corresponding to 135 ° still overlap, that is, as shown in fig. 6, the envelope surface images obtained in the shooting directions represented by the two parallel solid lines and the envelope surface images obtained in the shooting directions represented by the two dotted lines overlap mostly, specifically, the right side portions of the envelope surface images obtained in the shooting directions represented by the two parallel solid lines and the left side portions of the envelope surface images obtained in the shooting directions represented by the two dotted lines overlap.
Moreover, although the images of the two glass bulb surface images acquired at the adjacent shooting angles are overlapped, due to the difference of the visual angles, the images are distorted, and the position, the size and the shape of the real defect at the same position on the glass bulb in the two glass bulb surface images are different. However, although the position, size, shape of the real defect at the same position on the envelope may be different in both envelope surface images, the topological relationship between the real defects at the overlapping region is constant.
Therefore, the correct number of real defects can be obtained by calculating the corresponding matching relation of the real defects in the two glass bulb surface images acquired at the adjacent shooting angles. The calculation process of the corresponding matching relationship is as follows:
(1) a graphical structure representation of the surface image of each envelope is constructed.
Wherein, the node represents the real defect, and the edge represents the connecting line angle value between the central points of the real defect. And for each node, connecting four nodes closest to the node in the Euclidean distance to obtain edges, and forming a graph structure. An edge represents a direction value from the node to a neighboring node.
As shown in fig. 7, the representation of the nearest four nodes in the graph structure of the nodes a, b, c, d, e, and as shown in fig. 8, the representation of the neighboring four nodes of the node a when the node a is targeted.
The arrows in fig. 7 are bidirectional, indicating that two adjacent relations are mutual, i.e. b is one of the nearest four nodes of a, a is also one of the nearest four nodes of b, and the unidirectional arrow from a to d in fig. 7 indicates that d is one of the nearest four nodes of a, but a does not belong to one of the nearest four nodes of d.
(2) A representation of a node.
As shown in fig. 8, the nearest four-node information of each node can be obtained, the node is represented by edge information, each node has four edge information, a four-dimensional vector can be obtained, and the elements in the vector are sorted in ascending order to obtain the vector after ascending order.
For the node a in fig. 8, p (ab), p (ac), p (ae), p (ad) represent the weights of the four edges, and the ascending orders of [ p (ab), p (ac), p (ae), p (ad) ] are performed, and the ascending vector represents the topological relation of the node a.
(3) And calculating node matching.
The actual defects in the different images are first labeled.
And taking the cosine similarity of the vector representation of each node in the former graph structure and the vector representation of each node in the other graph structure as the adjacent structure similarity of the two nodes.
As shown in fig. 9, the left side node represents a node in the previous graph structure, the right side node represents a node in the subsequent graph structure, the node edge value represents the similarity of the neighboring structures, and the optimal matching result is obtained through optimal matching calculation, so as to obtain the corresponding matching relationship of the same real defect in different images.
In the embodiment, the connection of the glass bulb surface images at different shooting angles is completed by adopting the topological relation among the real defects in the glass bulb surface images at different shooting angles, and the connection of the glass bulb surface images at different shooting angles can be completed by adopting other conventional methods in other embodiments.
(4) A continuous matching chain is calculated.
As shown in fig. 10, the a and b real defects in the 0 ° captured glass bulb surface image correspond to the a1 and b1 real defects in the 45 ° captured glass bulb surface image, the a1 and b1 real defects in the 45 ° captured glass bulb surface image correspond to the a2 and b2 real defects in the 90 ° captured glass bulb surface image, and so on, the captured glass bulb surface images at all adjacent angles are matched, and a continuous matching chain of each real defect on the circumference of the glass bulb surface is calculated, for example: a-a1-a2-a3, is a true defect with a count of 1.
And taking the number of the finally obtained matching chains of the real defects as the total number of the real defects, and correspondingly determining the position and the size of each real defect on the circumference of the surface of the glass envelope.
6. And finishing the evaluation of the surface quality of the diode according to the number and the size of the real defects on the surface of the glass bulb.
The larger the number of real defects on the surface of the glass envelope, the larger the area ratio, and the poorer the surface quality of the glass envelope. From this the diode surface quality degradation value can be calculated:
wherein the content of the first and second substances,
denotes the number of real defects, S denotes the total area of real defects, and S denotes the total area of the envelope surface. q represents the diode surface quality degradation value, and the larger the q value is, the poorer the diode surface quality is.
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 (4)
1. A diode surface quality detection method is characterized by comprising the following steps:
vertically placing the transparent diode, horizontally irradiating the transparent diode by using a light source, rotating the transparent diode by using a set rotating angle as an interval, and shooting the complete circumference of the transparent diode by using a visible light means to obtain an even number of transparent diode glass bulb surface images and an even number of mapping images corresponding to the glass bulb surface images; the glass bulb surface image is shot at the same side of the light source, and the mapping image corresponding to the glass bulb surface image is shot at the opposite side of the light source;
performing linear detection on each glass bulb surface image, selecting any one pixel point on all determined straight lines as an initial point, determining a structural part area in the corresponding glass bulb surface image by a seed filling method, taking an area except the structural part area in the glass bulb surface image as a non-structural part area, and determining the structural part area and the non-structural part area in the corresponding mapping image according to a mapping relation; performing edge identification on the surface image of the glass envelope, removing structural part area outline edge information and glass envelope external outline edge information from the identified edge information, determining a defect area in the surface image of the glass envelope according to the residual edge information, and determining a defect area in a corresponding mapping image according to a mapping relation;
dividing two glass bulb surface images in the symmetrical direction and two corresponding mapping images into the same symmetrical group, wherein the shooting angles of the two glass bulb surface images in each symmetrical group are just opposite and the shooting angles of the two corresponding mapping images are just opposite, determining each overlapping area obtained by a defect area in the non-structural part area of the two mapping images in the symmetrical group, judging whether the ratio of the area of each overlapping area to the area of the two defect areas of the obtained overlapping area is larger than a set ratio, and if so, considering that the two defect areas of the obtained overlapping area are real defect areas; the two defect areas of the obtained overlapping area are defect areas corresponding to the positions of the overlapping area in the two mapping images; calculating the average gray value of the defect areas corresponding to the real defect area positions on the two glass bulb surface images in the symmetrical group, wherein the defect area with the smaller average gray value is the real defect in the non-structural part area on the glass bulb surface image;
determining whether the gray gradient direction of each defect area is vertical to the axis of the transparent diode glass bulb in the structural part area of the two glass bulb surface images of the symmetrical group, if not, the corresponding defect area is a real defect in the structural part area on the glass bulb surface image;
obtaining the real defects on the surface image of each glass bulb according to the real defects in the non-structural part area and the real defects in the structural part area on the surface image of each glass bulb, matching and connecting the surface images of the glass bulbs under any two adjacent shooting angles, and determining the number and the size of the real defects on the complete circumference of the transparent diode glass bulb;
and finishing the surface quality evaluation of the transparent diode according to the number and the size of the real defects on the complete circumference of the glass bulb.
2. The diode surface quality detection method according to claim 1, wherein the topological relation of each real defect in the graph structure of each glass bulb surface image is determined, and the matching and the connection of the glass bulb surface images at any two adjacent shooting angles are completed according to the topological relation of each real defect in the graph structure of each glass bulb surface image.
3. The diode surface quality detection method according to claim 1, wherein the specific method for completing the evaluation of the transparent diode surface quality by the number and size of real defects on the complete circumference of the glass bulb comprises the following steps:
calculating the surface quality degradation value of the diode:
wherein the content of the first and second substances,
representing the number of real defects, S representing the total area of the real defects, and S representing the total area of the surface of the glass envelope; the larger the q value, the lower the transparent diode surface quality evaluation.
4. The diode surface quality detection method of claim 1, wherein the set ratio is 90%, the set rotation angle is 45 °, and the number of the even number of transparent diode bulb surface images is 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210548200.5A CN115047006A (en) | 2022-05-20 | 2022-05-20 | Diode surface quality detection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210548200.5A CN115047006A (en) | 2022-05-20 | 2022-05-20 | Diode surface quality detection method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115047006A true CN115047006A (en) | 2022-09-13 |
Family
ID=83159818
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210548200.5A Pending CN115047006A (en) | 2022-05-20 | 2022-05-20 | Diode surface quality detection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115047006A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116561985A (en) * | 2023-04-18 | 2023-08-08 | 中国科学院国家空间科学中心 | Rapid evaluation method and system for quality reliability of diode |
-
2022
- 2022-05-20 CN CN202210548200.5A patent/CN115047006A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116561985A (en) * | 2023-04-18 | 2023-08-08 | 中国科学院国家空间科学中心 | Rapid evaluation method and system for quality reliability of diode |
| CN116561985B (en) * | 2023-04-18 | 2024-02-13 | 中国科学院国家空间科学中心 | Rapid evaluation method and system for quality reliability of diode |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109829914B (en) | Method and device for detecting product defects | |
| KR101951576B1 (en) | Inspection device and inspection method | |
| US6876445B2 (en) | Method for analyzing defect data and inspection apparatus and review system | |
| JP5723792B2 (en) | Image data processing method and system | |
| CN114723701A (en) | Gear defect detection method and system based on computer vision | |
| CN115020267B (en) | Semiconductor surface defect detection method | |
| BR112015001413B1 (en) | METHODS TO SEGMENT A TIRE SURFACE, AND TO DETECT DEFECTS ON A TIRE SURFACE, AND EQUIPMENT TO SEGMENT A TIRE SURFACE | |
| US20060007452A1 (en) | Coded target and photogrammetry method using such targets | |
| CN115047006A (en) | Diode surface quality detection method | |
| KR101782363B1 (en) | Vision inspection method based on learning data | |
| US20150302577A1 (en) | Method for the non-destructive testing of a blade preform | |
| CN109902576A (en) | A kind of training method and application of head shoulder images classifier | |
| CN113192013A (en) | Method and system for detecting defects of light-reflecting surface and electronic equipment | |
| CN114445348B (en) | New material water pump defect detection method and system based on optical means | |
| Park et al. | Automated defect inspection systems by pattern recognition | |
| CN107240109B (en) | Automatic detection method for instrument scale mark position | |
| JP3855360B2 (en) | Method and apparatus for inspecting streak defects | |
| JPH0546840A (en) | Coin discrimination device | |
| CN112036391A (en) | Electronic inspection method and system based on shale gas field production station | |
| CN116934752A (en) | Glass detection method and system based on artificial intelligence | |
| CN110199173A (en) | Position offset acquisition device, check device, position offset acquisition methods and inspection method | |
| CN114549513A (en) | Part identification method, part identification device, quality inspection method, electronic device, and storage medium | |
| CN113406112A (en) | Flaw detection method and system for transparent substrate film | |
| JPH03142576A (en) | Picture processing method | |
| CN115452845B (en) | LED screen surface damage detection method based on machine vision |
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 |