CN115272346A - PCB production process online detection method based on edge detection - Google Patents
PCB production process online detection method based on edge detection Download PDFInfo
- Publication number
- CN115272346A CN115272346A CN202211205424.2A CN202211205424A CN115272346A CN 115272346 A CN115272346 A CN 115272346A CN 202211205424 A CN202211205424 A CN 202211205424A CN 115272346 A CN115272346 A CN 115272346A
- Authority
- CN
- China
- Prior art keywords
- target image
- value
- gray
- threshold
- image block
- 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
- 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
- G06T7/11—Region-based segmentation
-
- 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
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
- G06T2207/30141—Printed circuit board [PCB]
Abstract
The invention discloses an edge detection-based PCB production process online detection method, which relates to the field of image processing and comprises the following steps: acquiring a gray image of a PCB to be detected; dividing the gray level image into a plurality of target image blocks; determining an initial threshold interval of each target image block according to the distance between a pixel point on a suspected circular hole edge line in each target image block and a circumferential center point and the gray distribution entropy, determining an initial threshold from the initial threshold interval, performing image enhancement on the target image blocks, obtaining an iteration threshold of each target image block, and determining a high threshold and a low threshold in each target image block according to the initial threshold and the iteration threshold of the target image blocks; according to the method and the device, the edge of the target image block is detected according to the high threshold and the low threshold of the target image block, and the defect area of the PCB to be detected is obtained.
Description
Technical Field
The invention relates to the field of image processing, in particular to an edge detection-based PCB production process online detection method.
Background
With the development of the electronic industry, PCB technology is applied to many fields such as computer hardware, home appliances, and medical devices. Due to different functional requirements of various industries, the PCB board patterns are more diversified and complex, and the more complex PCB boards are easy to generate errors of short circuit, solder bridge and open circuit in the production process, so that the quality detection needs to be carried out when the PCB boards leave a factory, and the problems of short circuit, solder bridge and open circuit in use are avoided.
The method for detecting the quality of the PCB at present is to determine whether the PCB has a defect area or not through template matching, the template matching needs to determine edge characteristics on the PCB to be detected firstly, and then the defect area is obtained according to characteristic points and the template matching, so accurate characteristic points need to be obtained if the detection result is accurate, but the edge detection of the PCB at present is common canny operator detection, because the surface circuit of the PCB is complex, and the gray difference of the defect place of a tiny round hole area is not obvious, the improper dual-threshold selection during the edge detection is easily caused, so the edge characteristics of the tiny round hole area on the surface of the PCB cannot be completely and accurately detected, if the round hole area has problems and is not detected, the subsequent use of the PCB is influenced, therefore, the invention provides an online detection method for the production process of the PCB based on the edge detection.
Disclosure of Invention
The invention provides an edge detection-based PCB production process online detection method, which aims to solve the existing problems.
The invention relates to an edge detection-based PCB production process online detection method, which adopts the following technical scheme:
acquiring a gray image of a PCB to be detected;
dividing the gray level image into a plurality of image blocks according to the initial size, and acquiring a plurality of target image blocks according to the difference between the mean value of the gray level information entropies of all the image blocks and the gray level information entropy of the gray level image;
acquiring a suspected round hole edge line in each target image block; obtaining a gray level distribution entropy of each pixel point according to the gray level value distribution probability and the gray level value of each pixel point in the gray level image; determining an initial threshold interval of each target image block according to the distance between the pixel point on the edge line of the suspected circular hole and the circumferential center point of the edge line of the suspected circular hole and the gray level distribution entropy of the pixel point on the edge line of the suspected circular hole and the circumferential center point;
acquiring an optimal segmentation threshold of an image formed by pixel points corresponding to the initial threshold interval of each target image block, and taking the optimal segmentation threshold obtained by each target image block as an initial threshold for iteration of the target image block;
carrying out image enhancement on each target image block in the gray level image to obtain an image-enhanced target image block and a gray level image; acquiring a gray value average value of suspected edge points in a gray image as an iteration threshold of each target image block, and if the difference value between the iteration threshold and the initial threshold of each target image block meets a set judgment threshold, taking the initial threshold of each target image block as a high threshold;
if the difference value between the iteration threshold value and the initial threshold value of the target image block does not accord with the set judgment threshold value, taking the iteration threshold value of the target image block as a new initial threshold value of the target image block, re-determining the new iteration threshold value of the target image block according to the mean value of the maximum gray value and the minimum gray value in the target image block, re-judging whether the difference value between the new initial threshold value and the new iteration threshold value of the target image block accords with the judgment threshold value or not, if not, continuing iteration until the difference value between the initial threshold value and the iteration threshold value of the target image block accords with the judgment threshold value, and taking the iteration threshold value obtained when the difference value accords with the judgment threshold value as a high threshold value of the corresponding target image block;
obtaining a low threshold value of a corresponding target image block according to the high threshold value of each target image block;
and performing edge detection on each target image block by using the high threshold and the low threshold of each target image block, and acquiring a defect area of the PCB to be detected according to an edge detection result.
Further, the step of obtaining a plurality of target image blocks according to the difference between the mean value of the grayscale information entropies of all the image blocks and the grayscale information entropies of the grayscale image includes:
if the difference value between the mean value of the gray scale information entropies of all the image blocks obtained for the first time and the gray scale information entropies of the gray scale images is smaller than or equal to a set threshold value, optionally dividing one image block obtained for the first time to obtain a plurality of image blocks with smaller sizes;
calculating the difference between the mean value of the gray scale information entropies of all the image blocks in the gray scale image and the gray scale information entropy of the gray scale image;
if the obtained difference is smaller than or equal to the set threshold, selecting one image block obtained for the first time to be divided to obtain a plurality of image blocks with smaller sizes;
calculating the difference value between the mean value of the gray scale information entropies of all image blocks in the gray scale image and the gray scale information entropy of the gray scale image;
if the obtained difference is less than or equal to the set threshold, continuing to divide the image block in the gray level image;
stopping image segmentation until the difference between the mean value of the gray scale information entropies of all the image blocks and the gray scale information entropies of the gray scale images is greater than a set threshold;
and recording all the divided image blocks in the gray-scale image as target image blocks.
Further, the step of obtaining the suspected round hole edge line in each target image block comprises:
carrying out non-maximum suppression on the gray level image, wherein the reserved points in the gray level image after the non-maximum suppression are suspected edge points;
carrying out Hough circle detection on all the suspected edge points to obtain suspected round hole edge lines in the gray level image;
and obtaining a suspected round hole edge line in the target image block according to the corresponding position of the target image block in the gray level image.
Further, a formula for obtaining the gray scale distribution entropy of each pixel point according to the gray scale value distribution probability and the gray scale value of each pixel point in the gray scale image is as follows:
wherein the content of the first and second substances,representing a gray value in a gray-scale image ofImage ofThe gray level distribution entropy of the pixel points;is shown asGray values of the class pixels;is shown asThe pixel points with the same gray value are the same type of pixel points;representing the total number of pixel points in the gray level image;representing a gray value in a gray image ofThe number of pixels of (2).
Further, the step of determining the initial threshold interval of each target image block according to the distance between the pixel point on the suspected circular hole edge line and the circumferential center point of the suspected circular hole edge line in each target image block and the gray level distribution entropy of the pixel point on the suspected circular hole edge line and the circumferential center point comprises the following steps:
calculating the gray scale dispersion of each pixel point on the edge line according to the distance between the pixel point on the edge line of the suspected circular hole in each target image block and the circumferential center point and the gray scale distribution entropy of the pixel point on the edge line of the suspected circular hole and the circumferential center point;
the formula for calculating the gray scale dispersion is as follows:
wherein, the first and the second end of the pipe are connected with each other,expressing the gray dispersion between the pixel point on the suspected round hole edge line and the circumference center point of the suspected round hole edge line;the coordinate on the edge line of the suspected round hole is represented asPixel point (c) and the circumferential center point of the edge line of the suspected round holeThe Euclidean distance between;the coordinate on the edge line of the suspected round hole is represented asThe gray level distribution entropy of the pixel points;expressing the gray distribution entropy of the circumferential center point of the suspected circular hole edge line;the coordinates of the circumference center point of the suspected round hole edge line are obtained;is a natural constant;
setting a boundary threshold, and acquiring the median of gray values of all pixel points of which the gray dispersion is greater than the boundary threshold in the target image block as an upper boundary value;
acquiring the median of the gray values of all the pixel points with the gray dispersion smaller than or equal to the boundary threshold value as a lower boundary value;
and obtaining an initial threshold interval of each target image block by using the upper boundary value and the lower boundary value of each target image block.
Further, the step of re-determining a new iteration threshold of the target image block according to the mean value of the maximum gray value and the minimum gray value in the target image block comprises:
obtaining the average value of the maximum gray value and the minimum gray value in the target image block;
Obtaining a mean value greater than in each target image blockMean value of gray values of pixel pointsAnd is equal to or less than the mean valueMean value of gray values of pixel points;
Two mean values obtained、Calculating the mean value as a new iteration threshold value of the target image block;
if the difference value between the new iteration threshold and the new initial threshold of the target image block does not accord with the set judgment threshold, acquiring the difference value in the target image blockToMean value of gray values of pixels betweenAndtoMean value of gray values of pixels between;
To the two obtained mean values、And calculating the mean value to be used as a new iteration threshold value of the target image block again.
Further, the formula for obtaining the low threshold value of each target image block according to the high threshold value of each target image block is as follows:
wherein the content of the first and second substances,a low threshold representing a target image block;a high threshold representing a target image block;judging a threshold value;to select the coefficients.
Further, the step of performing edge detection on each target image block by using the high threshold and the low threshold of each target image block to obtain the defect area of the PCB to be detected comprises the following steps:
judging whether two target image blocks need to be combined or not according to the high threshold and the low threshold of the two adjacent target image blocks, combining the target image blocks needing to be combined to obtain a target area, and obtaining the high threshold and the low threshold of the target area by using the high threshold and the low threshold of the two adjacent target image blocks;
and respectively carrying out edge detection on the corresponding target area and target image block by using the high threshold and the low threshold of all the merged target areas and the high threshold and the low threshold of the target image block without merging to obtain a defect area in the PCB to be detected.
Further, the step of judging whether the two target image blocks need to be combined according to the high threshold and the low threshold of the two adjacent target image blocks, and combining the target image blocks needing to be combined to obtain the target area comprises the following steps:
respectively calculating the proportion of gray values larger than a high threshold value in the suspected edge points of the two target image blocksRatio of gray value smaller than low threshold value in suspected edge pointAnd the proportion of the gray value of the suspected edge point being less than the high threshold value and greater than the low threshold value;
Interchanging the high threshold and the low threshold of two adjacent target image blocks, and respectively calculating the proportion of gray values larger than the high threshold in the suspected edge points in the target image blocks after interchanging the high threshold and the low thresholdRatio of gray value smaller than low threshold value in suspected edge pointThe gray value in the suspected edge point is less than the high threshold value and more than the high threshold valueLow threshold ratio;
Respectively calculating the proportion of two target image blocks after threshold value interchange、、And、、and if the difference obtained by the two target image blocks is less than or equal to the set threshold, combining the two adjacent target image blocks to obtain a target area.
Further, the step of obtaining the high threshold and the low threshold of the target area by using the high threshold and the low threshold of the two adjacent target image blocks includes:
taking the larger value of the two groups of high threshold values of the two adjacent target image blocks as the high threshold value of the target area;
and taking the smaller value of the two groups of low threshold values of the two adjacent target image blocks as the low threshold value of the target area.
The invention has the beneficial effects that: according to the PCB production process online detection method based on edge detection, the gray level image is segmented according to the difference between the gray level information entropies of all the image blocks and the gray level image to obtain the final target image block, so that the circular hole areas which are distributed intensively and have the same type are segmented into the same target image block to the greatest extent, and the optimal threshold value in each target image block obtained subsequently is more accurate; calculating the gray dispersion between the pixel points on the edge line of the suspected circular hole and the central point of the circumference to obtain suspected boundary points and suspected edge intermediate points, and limiting an interval for the subsequent selection of an initial threshold; the image is enhanced by utilizing logarithmic transformation, so that the initial threshold and the iteration threshold in the subsequent iteration process are more accurately calculated; the invention also obtains the double thresholds of each target image block in a self-adaptive manner according to an iterative algorithm, so that the edge characteristic result obtained by final edge detection is more accurate, and the defect detection result is more accurate.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in 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 for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a flowchart illustrating the general steps of an embodiment of an on-line detection method for PCB production process based on edge detection according to the present invention;
fig. 2 is a grayscale image of a PCB to be tested.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
As shown in fig. 1, an embodiment of the online detection method for the production process of the PCB based on edge detection of the present invention includes:
s1, obtaining a gray image of the PCB to be detected.
Specifically, an image of the PCB to be detected is collected, and the collected image is subjected to graying processing to obtain a grayscale image of the PCB to be detected.
In order to eliminate the influence of noise in the process of collecting images, the grayscale images are subjected to noise reduction processing.
It should be noted that, in the conventional method for reducing noise of an image, gaussian filtering is adopted, but the edges of the image become blurred in the denoising process by gaussian filtering, so that the bilateral filter replaces the gaussian filter, the bilateral filter considers the spatial domain information of the pixel points on the basis of filtering to increase the influence of the pixel gray value, when the gray value of the pixel points on the two sides of the edge has a large difference, the weight is reduced, the edge is prevented from being blurred, and more edge information in the image can be retained.
And S2, dividing the gray level image into a plurality of image blocks according to the initial size, and acquiring a plurality of target image blocks according to the difference between the mean value of the gray level information entropies of all the image blocks and the gray level information entropy of the gray level image.
The scheme is that edge detection optimization is carried out on a canny operator, noise reduction processing on a gray image is completed in the step S1, non-maximum suppression is carried out on the noise-reduced image subsequently, and then double-threshold selection is carried out.
It should be noted that the dual thresholds in the traditional canny operator need to be set by themselves, but the self-set dual thresholds are not suitable for the PCB image, because the PCB image is complex, the types of the circular areas in the image are not uniform, the circular areas with pins, the circular areas with holes, and the gray levels are different, so it is difficult to accurately select the thresholds, and therefore, in the scheme, the characteristic that the distribution of the circular hole areas on the PCB is locally concentrated is considered, as shown in fig. 2, the image is segmented, the circular holes which are distributed in a concentrated manner as much as possible are distributed into the same image block, and then the dual thresholds are selected for edge detection.
Specifically, the image segmentation is carried out on the grayscale image after noise reduction, the grayscale image is segmented into 4 image blocks with equal size, and the grayscale information entropy of each image block is calculated; and obtaining the mean value of the gray scale information entropies of all the image blocks, and comparing the obtained mean value with the gray scale information entropies of the gray scale images to evaluate the segmentation effect.
Calculating the gray level entropy value difference before and after the gray level image segmentation according to the following formula:
wherein the content of the first and second substances,expressing the gray level entropy value difference before and after the division, namely the difference between the mean value of the gray level information entropies of all the image blocks after the division and the gray level information entropy of the gray level image;a gray information entropy representing a gray image;and respectively representing the gray information entropy of each image block after division.
Setting a threshold valueUsed for judging whether to continuously divide the gray level image or not, and obtaining the gray level image according to the empirical valueIf the difference between the gray level entropy before and after the first divisionThe difference between the gray information entropy in the divided image block and the gray information entropy of the gray image is small, image division can be continuously carried out, and any one of the image blocks obtained by the first division is divided into an image with a smaller size of 4 equal divisions when the second division is carried out; if the difference between the gray level entropy before and after the first divisionExplaining that the difference between the gray information entropy in the image block and the gray information entropy of the gray image is larger, the image division is achievedAnd (4) for the purpose of cutting, the target image block is obtained without continuing to cut.
If the difference of the gray level entropy before and after the second divisionAnd selecting one of the image blocks obtained by the first division for division, and repeating the steps to obtain smaller image blocks by continuous division until the difference between the mean value of the gray scale information entropies of all the image blocks and the gray scale information entropies of the gray scale image is greater than a set threshold, stopping image division, marking all the divided image blocks in the gray scale image as target image blocks, and finally obtaining a plurality of target image blocks with different sizes.
It should be noted that the purpose of segmenting the grayscale image is to obtain the real distribution of the target area on the PCB, because the target area is not uniformly distributed on the PCB and is locally and intensively distributed, the target area refers to a circular hole area on the PCB, if image segmentation is not performed, a dual threshold is calculated from the entire grayscale image, which is more easily affected by the surface grayscale value, the grayscale value of the background pixel in the grayscale image is smaller than the grayscale value of the pixel in the target area, and the calculated low threshold may be smaller. The gray information entropy can reflect the complex situation of the image, for a pair of smooth and flat images with consistent colors, the gray value of each pixel point in the image is almost consistent, and the difference between the gray information entropy of the image after the image is segmented and the mean value of the gray information entropy of the image block after the image is segmented is very small; for an image with disordered gray value distribution, the gray information entropy value is large, if the gray value distribution in the image block after division is still disordered, and the gray information entropy of the image block is still large, the difference value of the gray entropy before and after division is not large, so that the image block needs to be continuously divided until the gray value distribution in the image block is uniform, namely the gray information entropy of the image block is small, the difference value of the gray entropy before and after division is large, and the difference value reaches a set threshold valueDescription of the concentrated distribution and the same speciesThe circular hole area is divided into the same image block, and the purpose of image segmentation is achieved.
S3, obtaining a suspected round hole edge line in each target image block; obtaining a gray value distribution entropy of each pixel point according to the gray value distribution probability and the gray value of each pixel point in the gray image; and determining an initial threshold interval of each target image block according to the distance between the pixel point on the edge line of the suspected circular hole and the circumferential center point of the edge line of the suspected circular hole and the gray distribution entropy of the pixel point on the edge line of the suspected circular hole and the circumferential center point.
Specifically, the gray level distribution entropy of each pixel point in the gray level image is calculated according to the following formula:
wherein the content of the first and second substances,representing a gray value in a gray image ofThe gray level distribution entropy of the pixel points;is shown asGray values of the class pixels;is shown asThe pixel points of the same gray value are the same type of pixel points;expressing the total number of pixel points in the gray level image;representing a gray value in a gray image ofThe number of pixel points of (a);representing a gray value in a gray image ofThe larger the proportion of the pixel points in the gray level image, namely the gray level value distribution probability of the pixel points, the more the distribution is expressed in the gray level image, and the reuse is carried outCalculating the reciprocal to show that the more the pixel points of the type are distributed in the gray level image, the smaller the obtained gray level distribution entropy of the pixel points is; then multiply byThe gray value of the similar pixel point is larger, and the gray distribution entropy of the obtained pixel point is larger.
It should be noted that, here, calculating the gray distribution entropy of the pixel point is to use the gray distribution entropy to represent the distribution probability of the gray value size and the gray value type of the pixel point, the first one isThe larger the gray value of the class pixel point is, the smaller the distribution range is, and the larger the obtained gray distribution entropy is.
The gray image is restrained and processed by using the non-maximum value, the reserved points in the gray image are suspected edge points and are not true edges, so that the true edges are obtained by thinning the edges according to the suspected edge points. The edge pixel point may be caused by stray effects, for example, color changes, and the edge pixel point caused by these color changes is located at the place where the circular hole area and the PCB board surface area are joined, that is, the boundary point of the circular hole.
And carrying out Hough circle detection on all the suspected edge points to obtain suspected round hole edge lines in the gray level image, and obtaining the suspected round hole edge lines in each target image block according to the corresponding positions of the target image blocks in the gray level image.
Calculating the gray dispersion between each pixel point on each suspected circular hole edge line in the target image block and the circumferential center point of the edge line according to the following formula:
wherein the content of the first and second substances,expressing the gray dispersion between the pixel point on the suspected round hole edge line and the circumference center point of the suspected round hole edge line;the coordinate on the edge line of the suspected round hole is represented asPixel point (c) and the circumference center point of the suspected round hole edge lineThe Euclidean distance between;the coordinate on the edge line of the suspected round hole is represented asThe gray level distribution entropy of the pixel points;expressing the gray distribution entropy of the circumferential center point of the suspected circular hole edge line;the coordinates of the circumference center point of the suspected round hole edge line are obtained;is a natural constant.
It should be noted that, the boundary point and the edge intermediate point on the edge line of the suspected circular hole are distinguished by two factors, on the PCB, the boundary point is distinguished from the edge intermediate point in that the closer the pixel point is to the boundary, the larger the gray value difference with the surrounding pixel points is, the larger the gray value difference with the circumferential center point is, the edge points are all around the edge intermediate point, and the gray value difference is relatively small; and the number of the boundary points is less than that of the intermediate points of the edge lines, and the distribution probability is smaller. And the Euclidean distance from the boundary point to the central point of the circumference is longer, and the Euclidean distance is usedThe larger the value is, the more likely the boundary point is, the distribution probability and the gray value size are expressed by the gray distribution entropy, the larger the gray distribution entropy is, the larger the gray value difference is, the smaller the gray distribution probability is, and the more likely the boundary point is. The suspected round hole edge line has a certain pixel width, and the edge middle point is the rest pixel points on the edge line within the value boundary point.
Setting a boundary threshold value to be 60 according to experience, and if the gray dispersion of pixel points on the suspected round hole edge line is greater than the boundary threshold value 60, determining the pixel points to be suspected boundary points; and if the gray dispersion of the pixel point on the edge line is less than or equal to the boundary threshold, the pixel point is a suspected edge intermediate point.
And acquiring a median value in the gray values of all the suspected boundary points in each target image block as an upper boundary value, acquiring a median value in the gray values of all the suspected edge intermediate points in each target image block as a lower boundary value, and acquiring an initial threshold interval, which is used for calculating an initial threshold, of the target image block subsequently by using the upper boundary value and the lower boundary value.
And S4, acquiring an optimal segmentation threshold of an image formed by pixel points corresponding to the initial threshold interval of each target image block, and taking the optimal segmentation threshold acquired by each target image block as an initial threshold for iteration of the target image block.
Specifically, an image composed of all pixel points corresponding to the gray value of an initial threshold interval in each target image block is obtained and recorded as a new image, each target image block corresponds to one new image, an optimal segmentation threshold of each new image is obtained by utilizing an Otsu algorithm, the specific mode of obtaining the optimal segmentation threshold is to arbitrarily select the segmentation threshold, divide the new image into a foreground part and a background part by utilizing the segmentation threshold, respectively obtain the proportion of the new image occupied by the pixel points of the foreground part and the background part, respectively obtain the gray value mean value of the pixel points of the foreground part and the background part, and calculate and obtain the between-class variance according to the proportion of the pixel points of the foreground part and the background part occupied by the new image and the gray value mean value; and adjusting the segmentation threshold to obtain different inter-class variances, wherein the segmentation threshold corresponding to the maximum inter-class variance is the optimal segmentation threshold, the Otsu algorithm is the prior art, and the optimal segmentation threshold is used as the initial threshold for subsequent iteration circulation.
Thus, an initial threshold value for each target image block to iterate is obtained.
S5, carrying out image enhancement on each target image block in the gray level image to obtain an image-enhanced target image block and a gray level image; acquiring a gray value average value of suspected edge points in a gray image as an iteration threshold of each target image block, and if the difference value between the iteration threshold and the initial threshold of the target image block conforms to a set judgment threshold, taking the initial threshold of the target image block as a high threshold; if the difference value between the iteration threshold value and the initial threshold value of the target image block does not accord with the set judgment threshold value, the iteration threshold value of the target image block is used as a new initial threshold value of the target image block, the new iteration threshold value of the target image block is determined again according to the mean value of the maximum gray value and the minimum gray value in the target image block, whether the difference value between the new initial threshold value and the new iteration threshold value of the target image block accords with the judgment threshold value or not is determined again, if not, iteration is continued until the difference value between the initial threshold value and the iteration threshold value of the target image block accords with the judgment threshold value, and the iteration threshold value obtained when the difference value accords with the judgment threshold value is used as the high threshold value of the corresponding target image block.
Specifically, each target image block is subjected to logarithmic transformation, and the contrast of a low-gray area in the image block is enhanced by the logarithmic transformation, namely the dark part details of the target image block are enhanced, so that the result obtained by subsequent processing is more accurate.
And obtaining a logarithmic curve of each target image block, obtaining a maximum gray value and a minimum gray value on the edge line of the suspected round hole in the target image block by using the logarithmic curve, wherein the point corresponding to the minimum gray value in the logarithmic curve is located at the position with the maximum slope of the curve, the point corresponding to the maximum gray value is located at the position with the minimum slope of the curve, and calculating the average value of the maximum gray value and the minimum gray value.
Using the mean of the maximum and minimum gray values of each target image blockClassifying the pixel points in the target image block, and enabling the pixel points to be larger than the mean valueThe pixel points are classified into one class, and are less than or equal to the mean valueThe pixel points of (a) are classified into one class.
Obtaining all the larger than average values in each target image blockThe gray value mean value of the pixel point is recorded as the first gray value mean value(ii) a Obtaining all average values less than or equal to in each target image blockThe gray value mean value of the pixel point is recorded as the second gray value mean value。
It should be noted that the mean value is calculatedFor obtaining a gray-scale division threshold value, first iterationA gray scale division threshold value, which changes in subsequent iterations; the gray value mean value of the two types of pixel points in the target image block is calculated, the distribution condition of the gray value in the target image block can be reflected, and therefore the obtained gray value mean value of the two types of pixel points can be used as an iteration threshold valueThe update parameter of (2).
And calculating the gray value average value of the pixel points in the gray image, wherein only the suspected edge points are restrained by the non-maximum value in the gray image at the moment, so that the obtained gray value average value is the gray value average value of all the suspected edge points, and the gray value average value of the gray image is used as the iteration threshold value of each target image block.
And judging the difference value between the iteration threshold value and the initial threshold value of each target image block according to the following formula:
wherein, the first and the second end of the pipe are connected with each other,an iteration threshold representing a target image block;an initial threshold value representing a target image block;taking the set judgment threshold value according to experience。
When the difference value between the initial threshold value and the iteration threshold value of the obtained target image blockThat is, when the difference is within the range of the judgment threshold, iteration is not required, and the iteration threshold is setAs the high threshold of the dual thresholds for edge detection subsequently performed on the target image block. In step S4, an initial threshold value is calculatedThe optimal segmentation threshold is obtained according to an Otsu algorithm, namely the initial threshold is used for distinguishing boundary points from edge intermediate points to the maximum extent so as to obtain more accurate and fine boundaries of the circular hole; whileThe gray value mean value of all the suspected edge points in the target image block, namely the gray value mean value of the boundary point and the edge intermediate point, reflects the gray distribution condition of the edge, so when the initial threshold value is usedAnd iteration thresholdThe smaller the difference therebetween, that is, the closer the iteration threshold value is to the optimal segmentation threshold value, the iteration threshold value at that time is considered to be capable of being used as a high threshold value to distinguish the boundary point and the edge intermediate point in the target image block.
When the target image block is obtainedIs compared with an iteration thresholdThat is, when the difference is outside the range of the judgment threshold, it is considered that the difference between the iteration threshold obtained at this time and the optimal segmentation threshold (initial threshold) is too large, and the boundary point cannot be accurately segmented, so that iteration is required to update the initial threshold and the iteration threshold.
The specific way of updating the initial threshold and the iteration threshold is to obtain the gray segmentation threshold of the initial iteration of each target image blockAnd the maximum gray value in the target image block is obtainedAnd minimum gray valueAnd in the target image blockMean value of gray values of corresponding pixelsAndmean value of gray values of corresponding pixels(ii) a The new iteration threshold of the second step target image blockIs updated toTaking the iteration threshold of the first step of the target image block as the target of the second stepNew initial threshold for the target image block, i.e.(ii) a Updating new initial threshold value according to target image blockAnd a new iteration thresholdComparing the difference with a judgment threshold, stopping iteration if the difference is less than or equal to the judgment threshold, and adding a new iteration thresholdHigh threshold as a subsequent edge detection of the target image block(ii) a If the obtained difference is larger than the judgment threshold, the iteration is continued, and the gray scale division threshold in the target image block is judged at the momentIs composed ofAndrespectively to obtain the mean values ofAndis the mean value ofAndof a third step of the target image blockNew iteration threshold is updated toTaking the iteration threshold of the second step of the target image block as a new initial threshold of the third stepI.e. the initial threshold is updated toCalculating the difference between the new initial threshold and the new iteration threshold in the third step, if the obtained difference does not conform to the set judgment threshold, sequentially iterating the target image block according to the iteration method until the difference between the iteration threshold and the initial threshold in a certain step is less than or equal to the judgment threshold, stopping iteration, and taking the iteration threshold of the target image block obtained at the moment as the high threshold for subsequent edge detection of the target image block。
It should be noted that the initial threshold value used for iterative determination at the beginning is an optimal segmentation threshold value obtained by using an Otsu algorithm, and divides the boundary points and the edge intermediate points to the greatest extent, and the initial iteration threshold value is an average value of the suspected edge points and represents the gray level distribution of the boundary points and the intermediate points on the edge line, so when the difference between the initial threshold value and the iteration threshold value meets a certain condition, it is considered that the true edge points with higher reliability can be obtained by the high threshold value obtained by iteration.
And determining a low threshold according to the high threshold of each target image block. Specifically, the low threshold of the target image block is calculated according to the following formula:
wherein the content of the first and second substances,representing blocks of an object imageA low threshold, specifically a low threshold of the two thresholds when the target image block is subjected to edge detection in the subsequent process;the high threshold value of the target image block is represented, and specifically is the high threshold value of the double threshold values when the target image block is subjected to edge detection in the subsequent process;judging a threshold value;for selecting coefficients, the determination is based on empirical values(ii) a Since the high threshold is about 2 times the low threshold in classical canny operator edge detection, the coefficients are taken(ii) a Using decision thresholdThe purpose of the subtraction is to avoid filtering out some real edge points with smaller gray values after the iteration is completed, because the smaller the lower threshold is, the more pixels are considered as real edge points after the upper and lower boundaries of the high and low thresholds are determined, and thus, the more edge information is retained.
Thus, a high threshold and a low threshold for each target image block are obtained.
And S6, performing edge detection on each target image block by using the high threshold and the low threshold of each target image block, and acquiring a defect area of the PCB to be detected according to an edge detection result.
Judging whether two target image blocks need to be combined or not according to the high threshold and the low threshold of the two adjacent target image blocks, combining the target image blocks needing to be combined to obtain a target area, obtaining the high threshold and the low threshold of the target area by using the high threshold and the low threshold of the two adjacent target image blocks, and respectively carrying out edge detection on the corresponding target area and the corresponding target image block by using the high threshold and the low threshold of all the combined target areas and the high threshold and the low threshold of the target image block which is not combined.
Specifically, taking the minimum target image block at the upper left corner in the grayscale image as an initial image block, respectively obtaining the dual thresholds of the target image block and the adjacent target image block, i.e. the high threshold and the low threshold of the target image block, and recording the dual thresholds of the target image block as the high threshold and the low threshold of the target image block、Recording the dual thresholds of the adjacent target image blocks of the target image block as、 Respectively calculating the gray value of the suspected edge point in each target image block to be greater than a high threshold valueIn a ratio ofThe gray value in the suspected edge point is less than the low threshold valueIn a ratio ofThe ratio of the gray value smaller than the high threshold value and larger than the low threshold value in the suspected edge points。
Interchanging the dual thresholds of the target image block and the adjacent target image block, and then according toTo obtain、、Respectively calculating the proportional relation between the suspected edge point and the high and low thresholds in each target image block, and sequentially recording the proportional relation as、、。
Respectively calculating three proportions of the target image block and each adjacent target image block after double-threshold interchange、、To correspond to、、The difference between the two, the sum of the differences beingIf it satisfiesIf the two thresholds of the target image block and the adjacent target image block have universality, the two target image blocks are combined to obtain a target area.
And taking the larger value of the two groups of high threshold values of the two target image blocks as the high threshold value of the target area, and taking the smaller value of the two groups of low threshold values of the two target image blocks as the low threshold value of the target area.
Traversing the whole image, and satisfying all interchange dual thresholdsThe adjacent target image blocks are combined to obtain a plurality of target areas, and a dual threshold of each target area is obtained, at this time, the combined target areas and the target image blocks which are not combined exist in the gray scale image.
Then interchanging double thresholds for adjacent target areas or adjacent target image blocks and target areas, and calculating the image after interchanging the double thresholdsWill all satisfyUntil there is no target area and target image block in the gray-scale image that can be combined.
Thus, all target regions, target image blocks, and dual thresholds for each target region and target image block in the grayscale image are obtained.
Reserving and marking the gray value as 255 for the pixel points with the gray value larger than the high threshold value in each target area \ target image block, marking the gray value of the pixel points with the gray value smaller than the low threshold value in the target area \ target image block as 0, extracting the pixel points with the gray value larger than the low threshold value and smaller than the high threshold value in the target area \ target image block, reserving and marking the gray value as 255 for the pixel points with the gray value larger than the high threshold value in the neighborhood of the pixel point, marking the gray value of the pixel point as 0 in the neighborhood of the pixel point as 0 in the case of not having the pixel points larger than the high threshold value, and marking all the gray values in the target area \ target image block as the edge pixel points in the target area \ target image block.
Respectively carrying out edge detection on the corresponding target area and target image block by utilizing the high threshold and the low threshold of all the merged target areas and the high threshold and the low threshold of the target image block which is not merged, optionally, splicing all the target areas and the target image block by double-threshold connection or boundary tracking (the double-threshold connection and the boundary tracking are the prior art), completing the edge detection of the circular hole area of the whole gray image, obtaining the edge characteristics of the gray image of the PCB to be detected, matching and comparing the PCB to be detected with the standard template to obtain the defect area of the PCB to be detected, and further judging whether the PCB meets the production standard, wherein the template matching is a known technology, and detailed description is omitted here.
In summary, the invention provides an edge detection-based online detection method for a production process of a PCB, which completes segmentation of a gray level image according to differences between gray level information entropies of all image blocks and the gray level image to obtain a final target image block, thereby maximally segmenting circular hole areas with the same type and concentrated distribution into the same target image block, and enabling an optimal threshold value in each subsequently obtained target image block to be more accurate; calculating the gray dispersion between pixel points on the edge line of the suspected round hole and the central point of the circumference to obtain boundary points and edge intermediate points, and defining an interval for the subsequent selection of an initial threshold; the image is enhanced by utilizing logarithmic transformation, so that the initial threshold and the iteration threshold in the subsequent iteration process are more accurately calculated; the invention also obtains the double thresholds of each target image block in a self-adaptive manner according to an iterative algorithm, so that the edge characteristic result obtained by final edge detection is more accurate, and the defect detection result is more accurate.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (10)
1. An edge detection-based PCB production process on-line detection method is characterized in that:
acquiring a gray image of a PCB to be detected;
dividing the gray level image into a plurality of image blocks according to the initial size, and acquiring a plurality of target image blocks according to the difference between the mean value of the gray level information entropies of all the image blocks and the gray level information entropy of the gray level image;
acquiring a suspected round hole edge line in each target image block; obtaining a gray level distribution entropy of each pixel point according to the gray level value distribution probability and the gray level value of each pixel point in the gray level image; determining an initial threshold interval of each target image block according to the distance between the pixel point on the edge line of the suspected round hole and the circumferential center point of the edge line of the suspected round hole and the gray level distribution entropy of the pixel point on the edge line of the suspected round hole and the circumferential center point;
acquiring an optimal segmentation threshold of an image formed by pixel points corresponding to the initial threshold interval of each target image block, and taking the optimal segmentation threshold obtained by each target image block as an initial threshold for iteration of the target image block;
carrying out image enhancement on each target image block in the gray level image to obtain an image-enhanced target image block and a gray level image; acquiring a gray value average value of suspected edge points in a gray image as an iteration threshold of each target image block, and if the difference value between the iteration threshold and the initial threshold of each target image block meets a set judgment threshold, taking the initial threshold of each target image block as a high threshold;
if the difference value between the iteration threshold value and the initial threshold value of the target image block does not accord with the set judgment threshold value, taking the iteration threshold value of the target image block as a new initial threshold value of the target image block, re-determining the new iteration threshold value of the target image block according to the mean value of the maximum gray value and the minimum gray value in the target image block, re-judging whether the difference value between the new initial threshold value and the new iteration threshold value of the target image block accords with the judgment threshold value or not, if not, continuing iteration until the difference value between the initial threshold value and the iteration threshold value of the target image block accords with the judgment threshold value, and taking the iteration threshold value obtained when the difference value accords with the judgment threshold value as a high threshold value of the corresponding target image block;
obtaining a low threshold value of a corresponding target image block according to the high threshold value of each target image block;
and performing edge detection on each target image block by using the high threshold and the low threshold of each target image block, and acquiring a defect area of the PCB to be detected according to an edge detection result.
2. The PCB production process on-line detection method based on edge detection as claimed in claim 1, wherein the step of obtaining a plurality of target image blocks according to the difference between the mean value of the gray scale information entropies of all the image blocks and the gray scale information entropy of the gray scale image comprises:
if the difference value between the mean value of the gray scale information entropies of all the image blocks obtained for the first time and the gray scale information entropies of the gray scale images is smaller than or equal to a set threshold value, optionally dividing one image block obtained for the first time to obtain a plurality of image blocks with smaller sizes;
calculating the difference value between the mean value of the gray scale information entropies of all the image blocks in the gray scale image and the gray scale information entropy of the gray scale image;
if the obtained difference is smaller than or equal to the set threshold, selecting one image block obtained for the first time to be divided to obtain a plurality of image blocks with smaller sizes;
calculating the difference between the mean value of the gray scale information entropies of all the image blocks in the gray scale image and the gray scale information entropy of the gray scale image;
if the obtained difference is less than or equal to the set threshold, continuing to divide the image block in the gray level image;
stopping image segmentation until the difference between the mean value of the gray scale information entropies of all the image blocks and the gray scale information entropies of the gray scale images is greater than a set threshold;
and recording all the divided image blocks in the gray-scale image as target image blocks.
3. The PCB production process on-line detection method based on edge detection as claimed in claim 1, wherein the step of obtaining the suspected round hole edge line in each target image block comprises:
carrying out non-maximum suppression on the gray level image, wherein the reserved points in the gray level image after the non-maximum suppression are suspected edge points;
carrying out Hough circle detection on all the suspected edge points to obtain suspected round hole edge lines in the gray level image;
and obtaining a suspected round hole edge line in the target image block according to the corresponding position of the target image block in the gray level image.
4. The PCB production process on-line detection method based on edge detection as claimed in claim 1, wherein the formula for obtaining the gray distribution entropy of each pixel point according to the gray value distribution probability and the gray value of each pixel point in the gray image is as follows:
wherein, the first and the second end of the pipe are connected with each other,representing a gray value in a gray-scale image ofThe gray level distribution entropy of the pixel points;is shown asGray values of the class pixel points;is shown asThe pixel points of the same gray value are the same type of pixel points;representing the total number of pixel points in the gray level image;representing a gray value in a gray-scale image ofThe number of pixels of (c).
5. The edge detection-based PCB production process online detection method of claim 1, wherein the step of determining the initial threshold interval of each target image block according to the distance between the pixel point on the suspected circular hole edge line and the circumferential center point of the suspected circular hole edge line and the gray scale distribution entropy of the pixel point on the suspected circular hole edge line and the circumferential center point comprises:
calculating the gray scale dispersion of each pixel point on the edge line according to the distance between the pixel point on the edge line of the suspected circular hole in each target image block and the circumferential center point and the gray scale distribution entropy of the pixel point on the edge line of the suspected circular hole and the circumferential center point;
the formula for calculating the gray scale dispersion is as follows:
wherein the content of the first and second substances,indicating a suspected round holeThe gray level dispersion between the pixel point on the edge line and the circumference center point of the suspected round hole edge line;the coordinate on the edge line of the suspected round hole is represented asPixel point (c) and the circumference center point of the suspected round hole edge lineThe Euclidean distance between;the coordinate on the edge line of the suspected round hole is represented asThe gray level distribution entropy of the pixel points;expressing the gray distribution entropy of the circumferential center point of the suspected circular hole edge line;the coordinates of the circumference center point of the suspected round hole edge line are obtained;is a natural constant;
setting a boundary threshold, and acquiring the median of gray values of all pixel points of which the gray dispersion is greater than the boundary threshold in the target image block as an upper boundary value;
acquiring the median of the gray values of all the pixel points with the gray dispersion smaller than or equal to the boundary threshold value as a lower boundary value;
and obtaining an initial threshold interval of each target image block by using the upper boundary value and the lower boundary value of each target image block.
6. The PCB production process on-line detection method based on edge detection as claimed in claim 1, wherein the step of re-determining the new iteration threshold of the target image block according to the mean of the maximum gray value and the minimum gray value in the target image block comprises:
obtaining the mean value of the maximum gray value and the minimum gray value in the target image block;
Obtaining a mean value greater than in each target image blockGray value mean value of pixel pointAnd is not more than the mean valueGray value mean value of pixel point;
Two mean values obtained、Calculating the mean value as a new iteration threshold value of the target image block;
if the difference value between the new iteration threshold and the new initial threshold of the target image block does not accord with the set judgment threshold, acquiring the difference value in the target image blockToMean value of gray values of pixels betweenAndtoMean value of gray values of pixels between;
7. The PCB production process on-line detection method based on edge detection as recited in claim 1, wherein the formula for obtaining the low threshold value of the corresponding target image block according to the high threshold value of each target image block is as follows:
8. The PCB production process on-line detection method based on edge detection as claimed in claim 1, wherein the step of performing edge detection on each target image block by using the high threshold and the low threshold of each target image block to obtain the defect area of the PCB to be detected comprises:
judging whether two target image blocks need to be combined or not according to the high threshold and the low threshold of the two adjacent target image blocks, combining the target image blocks needing to be combined to obtain a target area, and obtaining the high threshold and the low threshold of the target area by using the high threshold and the low threshold of the two adjacent target image blocks;
and respectively carrying out edge detection on the corresponding target area and target image block by using the high threshold and the low threshold of all the merged target areas and the high threshold and the low threshold of the target image block without merging to obtain a defect area in the PCB to be detected.
9. The PCB production process on-line detection method based on edge detection as recited in claim 8, wherein the step of judging whether the two target image blocks need to be merged according to the high threshold and the low threshold of the two adjacent target image blocks, and merging the target image blocks that need to be merged to obtain the target area comprises:
respectively calculating the proportion of gray values greater than a high threshold value in the suspected edge points of the two target image blocksAnd the ratio of the gray value of the suspected edge point to be less than the low threshold valueAnd the proportion of the gray value of the suspected edge point being less than the high threshold value and greater than the low threshold value;
Interchanging the high threshold and the low threshold of two adjacent target image blocks, and respectively calculating the proportion of gray values larger than the high threshold in the suspected edge points in the target image blocks after interchanging the high threshold and the low thresholdRatio of gray value smaller than low threshold value in suspected edge pointThe ratio of the gray value smaller than the high threshold value and larger than the low threshold value in the suspected edge points;
Respectively calculating the proportion of two target image blocks after threshold value interchange、、And、、difference between two target maps, ifAnd if the difference values obtained by the image blocks are less than or equal to the set threshold, combining two adjacent target image blocks to obtain a target area.
10. The PCB production process on-line detection method based on edge detection as claimed in claim 8, wherein the step of obtaining the high threshold and the low threshold of the target area by using the high threshold and the low threshold of two adjacent target image blocks comprises:
taking the larger value of the two groups of high threshold values of the two adjacent target image blocks as the high threshold value of the target area;
and taking the smaller value of the two groups of low threshold values of the two adjacent target image blocks as the low threshold value of the target area.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211205424.2A CN115272346A (en) | 2022-09-30 | 2022-09-30 | PCB production process online detection method based on edge detection |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211205424.2A CN115272346A (en) | 2022-09-30 | 2022-09-30 | PCB production process online detection method based on edge detection |
Publications (1)
Publication Number | Publication Date |
---|---|
CN115272346A true CN115272346A (en) | 2022-11-01 |
Family
ID=83757970
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202211205424.2A Pending CN115272346A (en) | 2022-09-30 | 2022-09-30 | PCB production process online detection method based on edge detection |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115272346A (en) |
Cited By (21)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115439474A (en) * | 2022-11-07 | 2022-12-06 | 山东天意机械股份有限公司 | Rapid positioning method for power equipment fault |
CN115457035A (en) * | 2022-11-10 | 2022-12-09 | 山东鲁旺机械设备有限公司 | Machine vision-based construction hanging basket welding quality detection method |
CN115546241A (en) * | 2022-12-06 | 2022-12-30 | 成都数之联科技股份有限公司 | Edge detection method, edge detection device, electronic equipment and computer readable storage medium |
CN115578383A (en) * | 2022-11-23 | 2023-01-06 | 惠州威尔高电子有限公司 | Thick copper PCB detection method based on panoramic image |
CN115689948A (en) * | 2023-01-05 | 2023-02-03 | 济宁智诚物业管理有限公司 | Image enhancement method for detecting cracks of building water supply pipeline |
CN115858832A (en) * | 2023-03-01 | 2023-03-28 | 天津市邱姆预应力钢绞线有限公司 | Method and system for storing production data of steel strand |
CN115908429A (en) * | 2023-03-08 | 2023-04-04 | 山东歆悦药业有限公司 | Foot bath powder grinding precision detection method and system |
CN116128877A (en) * | 2023-04-12 | 2023-05-16 | 山东鸿安食品科技有限公司 | Intelligent exhaust steam recovery monitoring system based on temperature detection |
CN116188462A (en) * | 2023-04-24 | 2023-05-30 | 深圳市翠绿贵金属材料科技有限公司 | Noble metal quality detection method and system based on visual identification |
CN116664453A (en) * | 2023-07-31 | 2023-08-29 | 山东中泳电子股份有限公司 | PET (polyethylene terephthalate) plate detection method for swimming touch plate |
CN116703921A (en) * | 2023-08-07 | 2023-09-05 | 东莞市溢信高电子科技有限公司 | Method for detecting quality of surface coating of flexible circuit board |
CN116758077A (en) * | 2023-08-18 | 2023-09-15 | 山东航宇游艇发展有限公司 | Online detection method and system for surface flatness of surfboard |
CN116883270A (en) * | 2023-07-04 | 2023-10-13 | 广州医科大学附属第四医院(广州市增城区人民医院) | Soft mirror clear imaging system for lithotripsy operation |
CN116883401A (en) * | 2023-09-07 | 2023-10-13 | 天津市生华厚德科技有限公司 | Industrial product production quality detection system |
CN116912277A (en) * | 2023-09-12 | 2023-10-20 | 山东鲁泰化学有限公司 | Circulating water descaling effect evaluation method and system |
CN117274722A (en) * | 2023-11-21 | 2023-12-22 | 深圳市咏华宇电子有限公司 | Intelligent detection method for distribution box based on infrared image |
CN117274291A (en) * | 2023-11-21 | 2023-12-22 | 深圳市京鼎工业技术股份有限公司 | Method for detecting mold demolding residues based on computer vision |
CN117392130A (en) * | 2023-12-12 | 2024-01-12 | 山东海纳智能装备科技股份有限公司 | On-line fault diagnosis system based on infrared image |
CN117437223A (en) * | 2023-12-20 | 2024-01-23 | 连兴旺电子(深圳)有限公司 | Intelligent defect detection method for high-speed board-to-board connector |
CN117455870A (en) * | 2023-10-30 | 2024-01-26 | 太康精密(中山)有限公司 | Connecting wire and connector quality visual detection method |
CN117522875A (en) * | 2024-01-08 | 2024-02-06 | 深圳市新创源精密智造有限公司 | Visual detection method for production quality of semiconductor carrier tape based on image filtering |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114155230A (en) * | 2021-12-07 | 2022-03-08 | 江苏普立特科技有限公司 | Quality classification method and system for injection molding PC board with smooth surface |
CN114862849A (en) * | 2022-07-06 | 2022-08-05 | 山东智领新材料有限公司 | Aluminum alloy plate film coating effect evaluation method based on image processing |
CN115049653A (en) * | 2022-08-15 | 2022-09-13 | 凤芯微电子科技(聊城)有限公司 | Integrated circuit board quality detection system based on computer vision |
CN115063400A (en) * | 2022-07-22 | 2022-09-16 | 山东中艺音美器材有限公司 | Musical instrument production defect detection method using visual means |
-
2022
- 2022-09-30 CN CN202211205424.2A patent/CN115272346A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114155230A (en) * | 2021-12-07 | 2022-03-08 | 江苏普立特科技有限公司 | Quality classification method and system for injection molding PC board with smooth surface |
CN114862849A (en) * | 2022-07-06 | 2022-08-05 | 山东智领新材料有限公司 | Aluminum alloy plate film coating effect evaluation method based on image processing |
CN115063400A (en) * | 2022-07-22 | 2022-09-16 | 山东中艺音美器材有限公司 | Musical instrument production defect detection method using visual means |
CN115049653A (en) * | 2022-08-15 | 2022-09-13 | 凤芯微电子科技(聊城)有限公司 | Integrated circuit board quality detection system based on computer vision |
Cited By (38)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115439474A (en) * | 2022-11-07 | 2022-12-06 | 山东天意机械股份有限公司 | Rapid positioning method for power equipment fault |
CN115457035A (en) * | 2022-11-10 | 2022-12-09 | 山东鲁旺机械设备有限公司 | Machine vision-based construction hanging basket welding quality detection method |
CN115457035B (en) * | 2022-11-10 | 2023-03-24 | 山东鲁旺机械设备有限公司 | Machine vision-based construction hanging basket welding quality detection method |
CN115578383A (en) * | 2022-11-23 | 2023-01-06 | 惠州威尔高电子有限公司 | Thick copper PCB detection method based on panoramic image |
CN115578383B (en) * | 2022-11-23 | 2023-04-07 | 惠州威尔高电子有限公司 | Thick copper PCB detection method based on panoramic image |
CN115546241A (en) * | 2022-12-06 | 2022-12-30 | 成都数之联科技股份有限公司 | Edge detection method, edge detection device, electronic equipment and computer readable storage medium |
CN115689948A (en) * | 2023-01-05 | 2023-02-03 | 济宁智诚物业管理有限公司 | Image enhancement method for detecting cracks of building water supply pipeline |
CN115858832A (en) * | 2023-03-01 | 2023-03-28 | 天津市邱姆预应力钢绞线有限公司 | Method and system for storing production data of steel strand |
CN115858832B (en) * | 2023-03-01 | 2023-05-02 | 天津市邱姆预应力钢绞线有限公司 | Method and system for storing production data of steel strand |
CN115908429B (en) * | 2023-03-08 | 2023-05-19 | 山东歆悦药业有限公司 | Method and system for detecting grinding precision of foot soaking powder |
CN115908429A (en) * | 2023-03-08 | 2023-04-04 | 山东歆悦药业有限公司 | Foot bath powder grinding precision detection method and system |
CN116128877A (en) * | 2023-04-12 | 2023-05-16 | 山东鸿安食品科技有限公司 | Intelligent exhaust steam recovery monitoring system based on temperature detection |
CN116188462A (en) * | 2023-04-24 | 2023-05-30 | 深圳市翠绿贵金属材料科技有限公司 | Noble metal quality detection method and system based on visual identification |
CN116188462B (en) * | 2023-04-24 | 2023-08-11 | 深圳市翠绿贵金属材料科技有限公司 | Noble metal quality detection method and system based on visual identification |
CN116883270A (en) * | 2023-07-04 | 2023-10-13 | 广州医科大学附属第四医院(广州市增城区人民医院) | Soft mirror clear imaging system for lithotripsy operation |
CN116883270B (en) * | 2023-07-04 | 2024-03-22 | 广州医科大学附属第四医院(广州市增城区人民医院) | Soft mirror clear imaging system for lithotripsy operation |
CN116664453A (en) * | 2023-07-31 | 2023-08-29 | 山东中泳电子股份有限公司 | PET (polyethylene terephthalate) plate detection method for swimming touch plate |
CN116664453B (en) * | 2023-07-31 | 2023-10-20 | 山东中泳电子股份有限公司 | PET (polyethylene terephthalate) plate detection method for swimming touch plate |
CN116703921A (en) * | 2023-08-07 | 2023-09-05 | 东莞市溢信高电子科技有限公司 | Method for detecting quality of surface coating of flexible circuit board |
CN116703921B (en) * | 2023-08-07 | 2023-12-05 | 东莞市溢信高电子科技有限公司 | Method for detecting quality of surface coating of flexible circuit board |
CN116758077A (en) * | 2023-08-18 | 2023-09-15 | 山东航宇游艇发展有限公司 | Online detection method and system for surface flatness of surfboard |
CN116758077B (en) * | 2023-08-18 | 2023-10-20 | 山东航宇游艇发展有限公司 | Online detection method and system for surface flatness of surfboard |
CN116883401A (en) * | 2023-09-07 | 2023-10-13 | 天津市生华厚德科技有限公司 | Industrial product production quality detection system |
CN116883401B (en) * | 2023-09-07 | 2023-11-10 | 天津市生华厚德科技有限公司 | Industrial product production quality detection system |
CN116912277B (en) * | 2023-09-12 | 2023-12-12 | 山东鲁泰化学有限公司 | Circulating water descaling effect evaluation method and system |
CN116912277A (en) * | 2023-09-12 | 2023-10-20 | 山东鲁泰化学有限公司 | Circulating water descaling effect evaluation method and system |
CN117455870B (en) * | 2023-10-30 | 2024-04-16 | 太康精密(中山)有限公司 | Connecting wire and connector quality visual detection method |
CN117455870A (en) * | 2023-10-30 | 2024-01-26 | 太康精密(中山)有限公司 | Connecting wire and connector quality visual detection method |
CN117274722B (en) * | 2023-11-21 | 2024-01-26 | 深圳市咏华宇电子有限公司 | Intelligent detection method for distribution box based on infrared image |
CN117274291B (en) * | 2023-11-21 | 2024-02-13 | 深圳市京鼎工业技术股份有限公司 | Method for detecting mold demolding residues based on computer vision |
CN117274291A (en) * | 2023-11-21 | 2023-12-22 | 深圳市京鼎工业技术股份有限公司 | Method for detecting mold demolding residues based on computer vision |
CN117274722A (en) * | 2023-11-21 | 2023-12-22 | 深圳市咏华宇电子有限公司 | Intelligent detection method for distribution box based on infrared image |
CN117392130A (en) * | 2023-12-12 | 2024-01-12 | 山东海纳智能装备科技股份有限公司 | On-line fault diagnosis system based on infrared image |
CN117392130B (en) * | 2023-12-12 | 2024-02-23 | 山东海纳智能装备科技股份有限公司 | On-line fault diagnosis system based on infrared image |
CN117437223A (en) * | 2023-12-20 | 2024-01-23 | 连兴旺电子(深圳)有限公司 | Intelligent defect detection method for high-speed board-to-board connector |
CN117437223B (en) * | 2023-12-20 | 2024-02-23 | 连兴旺电子(深圳)有限公司 | Intelligent defect detection method for high-speed board-to-board connector |
CN117522875A (en) * | 2024-01-08 | 2024-02-06 | 深圳市新创源精密智造有限公司 | Visual detection method for production quality of semiconductor carrier tape based on image filtering |
CN117522875B (en) * | 2024-01-08 | 2024-04-05 | 深圳市新创源精密智造有限公司 | Visual detection method for production quality of semiconductor carrier tape based on image filtering |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN115272346A (en) | PCB production process online detection method based on edge detection | |
CN116168026B (en) | Water quality detection method and system based on computer vision | |
CN114418957B (en) | Global and local binary pattern image crack segmentation method based on robot vision | |
CN109242853B (en) | PCB defect intelligent detection method based on image processing | |
WO2017121018A1 (en) | Method and apparatus for processing two-dimensional code image, and terminal and storage medium | |
CN111583216A (en) | Defect detection method for PCBA | |
CN111598801B (en) | Identification method for weak Mura defect | |
JP2024050880A (en) | Character segmentation method, device, and computer-readable storage medium | |
CN111709964B (en) | PCBA target edge detection method | |
CN110378893B (en) | Image quality evaluation method and device and electronic equipment | |
CN115775250A (en) | Golden finger circuit board defect rapid detection system based on digital image analysis | |
CN116735612B (en) | Welding defect detection method for precise electronic components | |
CN114332026A (en) | Visual detection method and device for scratch defects on surface of nameplate | |
CN116542982A (en) | Departure judgment device defect detection method and device based on machine vision | |
CN117094975A (en) | Method and device for detecting surface defects of steel and electronic equipment | |
TW202127372A (en) | Method for defect level determination and computer readable storage medium thereof | |
CN112419207A (en) | Image correction method, device and system | |
CN115439523A (en) | Method and equipment for detecting pin size of semiconductor device and storage medium | |
CN115272350A (en) | Method for detecting production quality of computer PCB mainboard | |
CN114998311A (en) | Part precision detection method based on homomorphic filtering | |
CN117152165B (en) | Photosensitive chip defect detection method and device, storage medium and electronic equipment | |
CN112819827B (en) | LED electrode offset detection method and device based on polar coordinate transformation and storage medium | |
CN116883408B (en) | Integrating instrument shell defect detection method based on artificial intelligence | |
CN115254674B (en) | Bearing defect sorting method | |
CN109211919B (en) | Method and device for identifying magnetic tile defect area |
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 |