CN113588655A - Detection device for surface defects of MDF fiber lines and working method thereof - Google Patents
Detection device for surface defects of MDF fiber lines and working method thereof Download PDFInfo
- Publication number
- CN113588655A CN113588655A CN202110731154.8A CN202110731154A CN113588655A CN 113588655 A CN113588655 A CN 113588655A CN 202110731154 A CN202110731154 A CN 202110731154A CN 113588655 A CN113588655 A CN 113588655A
- Authority
- CN
- China
- Prior art keywords
- image
- mdf fiber
- camera
- gradient
- mdf
- 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.)
- Withdrawn
Links
- 239000000835 fiber Substances 0.000 title claims abstract description 95
- 230000007547 defect Effects 0.000 title claims abstract description 85
- 238000001514 detection method Methods 0.000 title claims abstract description 53
- 238000000034 method Methods 0.000 title claims abstract description 31
- 238000004364 calculation method Methods 0.000 claims description 14
- 238000005516 engineering process Methods 0.000 claims description 14
- 238000012937 correction Methods 0.000 claims description 11
- 238000013136 deep learning model Methods 0.000 claims description 10
- 238000003708 edge detection Methods 0.000 claims description 8
- 238000005070 sampling Methods 0.000 claims description 4
- 230000002950 deficient Effects 0.000 claims description 3
- 238000009792 diffusion process Methods 0.000 claims description 3
- 230000005764 inhibitory process Effects 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 238000013461 design Methods 0.000 abstract description 5
- 230000000007 visual effect Effects 0.000 abstract description 3
- 238000007689 inspection Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 230000007246 mechanism Effects 0.000 description 3
- 238000007781 pre-processing Methods 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000005286 illumination Methods 0.000 description 2
- 238000010191 image analysis Methods 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 230000001788 irregular Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 208000034656 Contusions Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 239000011521 glass Substances 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- 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
- G01N2021/8887—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 based on image processing techniques
Abstract
A detection device for MDF fiber line surface defects and a working method thereof comprise a light source, a camera component, a camera bracket and an industrial personal computer; the light source comprises an annular LED positive light source and a rectangular LED auxiliary light source, the camera component comprises a CCD camera, a lens and an image acquisition card, and the camera support is of a double-column gantry structure and is arranged above a conveyor for conveying MDF fiber lines to perform surface defect detection; the annular LED positive light source is arranged right above the camera support, and the rectangular LED auxiliary light sources are arranged in the upper middle of the cylindrical surface of each upright column of the camera support, which faces one side of the conveyor; the CCD camera is arranged in the middle of the bottom surface of the camera support beam. The device for detecting the surface defects of the MDF fiber lines and the working method thereof have reasonable structural design, adopt non-contact visual detection to carry out all-around defect detection on the surface of the MDF fiber lines, and have simple working method, high precision, high efficiency and high intelligent degree.
Description
Technical Field
The invention belongs to the technical field of detection devices, and particularly relates to a detection device for surface defects of MDF (medium density fiber) lines and a working method thereof.
Background
In order to guarantee the quality of the MDF fiber line products, enterprises in China generally adopt a quality inspector sampling inspection mode to carry out quality inspection on the MDF fiber line products. The manual spot inspection has the defects of missing inspection, false inspection, low efficiency and the like. Along with the increasing standardization and strict control of the national product standard, the price paid by enterprises due to the product quality problem is increased continuously, the labor cost is increased year by year, and the enterprises urgently need to realize transformation and upgrading from manual spot inspection to automatic and intelligent detection.
The surface defect detection is an extremely important part in the industrial production of the MDF fiber line products, and the accuracy of the surface defect detection can directly influence the quality of the MDF fiber line products. Therefore, the core technology such as design, manufacture, detection algorithm and the like of the detection device for the surface defects of the MDF fiber line products is particularly important. Therefore, there is a need to develop an apparatus for detecting surface defects of MDF fiber strands and a method for operating the same.
Chinese patent application No. CN201110077446.0 discloses a glass bottle defect detecting platform, detection mechanism include bottleneck detection device, bottle end detection device, bottle shoulder detection device and body detection device, detection mechanism is connected with the main control system, is not designed to MDF fiber line product, does not make improvement and improvement in intelligence, efficiency, accuracy, response speed of detection mechanism.
Disclosure of Invention
The purpose of the invention is as follows: in order to overcome the defects, the invention aims to provide the device for detecting the surface defects of the MDF fiber lines and the working method thereof, the device is reasonable in structural design, adopts non-contact visual detection to carry out all-round defect detection on the surface of the MDF fiber lines, and is simple in working method, high in precision and efficiency, high in intelligent degree and wide in application prospect.
The purpose of the invention is realized by the following technical scheme:
a device for detecting surface defects of MDF fiber lines comprises a light source, a camera assembly, a camera bracket and an industrial personal computer; the light source comprises an annular LED positive light source and a rectangular LED auxiliary light source, the camera component comprises a CCD camera, a lens and an image acquisition card, and the camera support is of a double-column gantry structure and is arranged above a conveyor for conveying MDF fiber lines to perform surface defect detection; the annular LED positive light source is arranged right above the camera support, and the rectangular LED auxiliary light sources are arranged in the upper middle of the cylindrical surface of each upright column of the camera support, which faces one side of the conveyor; the CCD camera is arranged in the middle of the bottom surface of the camera bracket beam; the CCD camera is connected with an image acquisition card, a lens is arranged at the bottom of the CCD camera, and the image acquisition card transmits the shot image data to the industrial personal computer in real time through the network port; the industrial personal computer preprocesses image data in the 5G edge calculation layer, uploads the image data to the cloud for calculation, and finally learns multiple defect modes in the PasS layer by combining a surface defect model, a point cloud technology and a deep learning model of the MDF fiber line, so that the surface defect of the MDF fiber line is accurately detected.
The device for detecting the surface defects of the MDF fiber lines has reasonable structural design, the CCD camera has the advantages of high imaging quality and high-speed imaging, and is convenient to detect the MDF fiber lines moving on the conveyor, the CCD camera is matched with a lens to capture images on the surfaces of the MDF fiber lines, the acquired images are transmitted to an industrial personal computer through an image acquisition card to realize data acquisition and pretreatment by adopting a 5G edge calculation layer, and by utilizing a 5G transmission technology, each device has the advantages of high speed, low delay, low energy, low cost and the like in the process of acquiring and transmitting image data; through deep research on the surface defect generation process and characteristics of the MDF fiber line product, the PaaS layer extracts defect common characteristics such as burrs, bruises, appearance color differences, size errors and the like from 3 aspects of image preprocessing, defect detection and positioning, feature extraction and classification through multi-technology fusion such as big data, cloud computing, image recognition, 3D point cloud, CCD and the like, and the surface microscopic defect detection of the MDF fiber line product is constructed on the basis of the bottom layer capability of an internet platform, so that the accurate judgment on the surface defects of the MDF fiber line is achieved.
Further, according to the device for detecting the surface defects of the MDF fiber lines, a photoelectric sensor is arranged on the lower portion, facing the cylindrical surface of one side of the conveyor, of each stand column of the camera support, and the photoelectric sensor is connected with the industrial personal computer through a serial port.
Further, according to the device for detecting the surface defects of the MDF fiber lines, the CCD camera adopts a high-resolution industrial digital CCD camera, and the lens adopts a double telecentric machine vision lens.
The invention uses the high-resolution industrial digital CCD camera in cooperation with the vision lens of the double telecentric machine, and the shot and displayed picture has excellent resolution.
Further, the working method of the device for detecting the surface defects of the MDF fiber lines comprises the following steps:
(1) placing an MDF fiber line to be subjected to surface defect detection on a conveyor, starting the conveyor, and when the MDF fiber line reaches the position right below a CCD camera, transmitting an arrival signal to an industrial personal computer by a photoelectric sensor, controlling the conveyor to stop running and sending a CCD camera photographing signal to the industrial personal computer;
(2) the CCD camera takes a picture and transmits the acquired image to the industrial personal computer through the image acquisition card;
(3) the industrial personal computer preprocesses the acquired image in a 5G edge calculation layer, uploads the image to the cloud for calculation, and finally learns a plurality of defect modes in the PasS layer by combining a surface defect model, a point cloud technology and a deep learning model of the MDF fiber line, so that the surface defect of the MDF fiber line can be accurately detected; when the industrial personal computer detects that the MDF fiber line has defects, the industrial personal computer drives the seven-degree-of-freedom heavy-load mechanical arm to grab the defective MDF fiber line to the unqualified product area.
Further, the working method of the device for detecting the surface defects of the MDF fiber strands comprises the following steps:
(1) graying: converting the color image into a gray image by adopting a weighted average gray method for the acquired image;
(2) and (3) straight line detection: carrying out linear detection on the grayed image by an LSD linear detection method;
(3) and (3) inclination correction: screening the image after the straight line detection according to conditions, acquiring a long edge of an MDF fiber line in the image, calculating a difference value of coordinates at two ends of the long edge in the horizontal and vertical directions, and obtaining a slope k and an inclination angle theta of the straight line where the long edge is located, so as to perform image inclination correction;
(4) edge detection: removing noise of the image after inclination correction by adopting a Gaussian filter, convolving each pixel point in the image by a Gaussian kernel, replacing the value of a central pixel point with the weighted average gray value of the pixel in the region, respectively calculating the amplitude values of the pixel points in the x direction and the y direction by adopting a Sobel operator, convolving the image with the kernel of the size of 3 multiplied by 3, calculating the gradient in the horizontal direction and the vertical direction, and solving the approximate gradient and the gradient direction; a non-maximum value inhibition method is adopted for the gradient amplitude, a local maximum value is searched, and non-edge pixels are eliminated redundantly; comparing and judging the amplitude of each pixel point with a set hysteresis threshold, wherein the pixel point is an edge pixel when the amplitude of the pixel point is greater than a high threshold, and is excluded when the amplitude of the pixel point is less than a low threshold, and if the amplitude is between the high threshold and the low threshold, the pixel is reserved only when the pixel is adjacent to a pixel which is higher than the high threshold;
(5) template matching: and (3) learning various defect modes by combining the image subjected to edge detection with a surface defect model, a point cloud technology and a deep learning model of the MDF fiber line in the PasS layer, and judging the surface defects of the MDF fiber line.
After the industrial personal computer obtains the image of the MDF fiber line to be measured, because in the image acquisition process, the phenomena of poor illumination, irregular operation and the like can be inevitably generated, the obtained image has deviation from the ideal, such as noise, the position of the image needs to be corrected and the like, and the problems can cause great interference on the image analysis. In order to better identify the characteristic information to be detected in the image of the MDF fiber line, the image needs to be preprocessed first, so that the useful information of the MDF fiber line is enhanced, and the subsequent defect identification and detection on the MDF fiber line are faster and more accurate.
The preprocessing steps are reasonably designed, and because the detection and identification of the surface defects of the MDF fiber lines are irrelevant to the color characteristics of the MDF fiber lines, the color images are converted into gray level images, so that the algorithm processing speed can be increased; in order to avoid inclination of the positions of the acquired MDF fiber lines of the pictures in different degrees and improve the accuracy of subsequent detection, the inclination correction of the images is realized according to the slope of a known horizontal straight line by searching for the known horizontal straight line; the edge features are used as important matching features, and the accuracy of matching between the picture and the surface defect model of the MDF fiber line can be improved through edge detection.
Further, the working method of the device for detecting the surface defects of the MDF fiber lines comprises the following steps:
(1) scaling the image to 80% of the original image through Gaussian down-sampling;
(2) calculating the gradient of each pixel point in the horizontal direction and the gradient of each pixel point in the vertical direction, solving a gradient angle according to the horizontal gradient and the vertical gradient, and calculating a gradient amplitude;
(3) setting the gradient value of each pixel point of the image between 0 and 1023, and creating 1024 linked lists; through gradient traversal of pixel points, pixel point coordinates with the same gradient value are placed into the same linked list, each linked list is sorted according to the sequence from large to small, and finally the head and the tail of each linked list are linked to form a large linked list;
(4) traversing the whole linked list, taking the coordinate position of a pixel point stored at the head of the linked list as a seed pixel, diffusing the seed pixel through a region growing algorithm according to the direction with similar gradient, and removing the diffused pixel point from the linked list;
(5) performing rectangle fitting on the pixel region obtained by diffusion;
(6) calculating the precision error of the fitted rectangle according to an NFA formula, and if the precision error meets the requirement, considering the rectangle as an effective straight line; and repeating the step 4) until the traversal of the whole linked list is completed.
Compared with the prior art, the invention has the following beneficial effects:
(1) the structure is reasonable in design, the MDF fiber line surface is subjected to all-around defect detection by adopting non-contact visual detection, the working method is simple, the precision and the efficiency are high, the intelligent degree is high, and the application prospect is wide;
(2) the CD camera is matched with a lens to capture images of the surface of the MDF fiber line, the acquired images are transmitted to an industrial personal computer through an image acquisition card to realize data acquisition and pretreatment by adopting a 5G edge calculation layer, and by utilizing a 5G transmission technology, each device has the advantages of high speed, low delay, low energy, low cost and the like in the process of acquiring and transmitting image data; through deep research on the surface defect generation process and characteristics of the MDF fiber line product, a surface defect model, a point cloud technology and a deep learning model of the MDF fiber line are combined in the PasS layer to learn various defect modes, and accurate detection of the surface defects of the MDF fiber line is achieved.
Drawings
FIG. 1 is a layout diagram of the device for detecting the surface defects of the MDF fiber lines;
FIG. 2 is a structural diagram of the device for detecting the surface defects of the MDF fiber lines;
in the figure: the device comprises a light source 1, an annular LED positive light source 11, a rectangular LED auxiliary light source 12, a camera component 2, a CCD (charge coupled device) camera 21, a lens 22, an image acquisition card 23, a camera support 3, an industrial personal computer 4, a photoelectric sensor 5 and a conveyor a.
Detailed Description
The technical solutions in the embodiments of the present invention will be described clearly and completely with reference to the following specific embodiments and accompanying fig. 1-2, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all 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, the following embodiments provide an apparatus for detecting surface defects of an MDF fiber strand, including a light source 1, a camera assembly 2, a camera support 3, and an industrial personal computer 4; the light source 1 comprises an annular LED positive light source 11 and a rectangular LED auxiliary light source 12, the camera component 2 comprises a CCD camera 21, a lens 22 and an image acquisition card 23, and the camera support 3 is of a double-column gantry structure and is arranged above a conveyor for conveying MDF fiber lines to perform surface defect detection; the annular LED positive light source 11 is arranged right above the camera support 3, and the rectangular LED auxiliary light sources 12 are arranged in the upper middle of the cylindrical surface of each upright column of the camera support 3 facing one side of the conveyor; the CCD camera 21 is arranged in the middle of the bottom surface of the beam of the camera bracket 3; the CCD camera 21 is connected with an image acquisition card 23, a lens 22 is installed at the bottom of the CCD camera 21, and the image acquisition card 23 transmits shot image data to the industrial personal computer 2 in real time through a network port; the industrial personal computer 4 preprocesses image data in the 5G edge calculation layer, uploads the image data to the cloud for calculation, and finally learns multiple defect modes in the PasS layer by combining a surface defect model, a point cloud technology and a deep learning model of the MDF fiber line, so that the surface defect of the MDF fiber line is accurately detected.
Furthermore, a photoelectric sensor 5 is arranged on the lower portion of the cylindrical surface of each upright column of the camera support 3, which faces one side of the conveyor, and the photoelectric sensor 5 is connected with an industrial personal computer 4 through a serial port.
Examples
The working method of the device for detecting the surface defects of the MDF fiber lines comprises the following steps:
(1) placing an MDF fiber line to be subjected to surface defect detection on a conveyor, starting the conveyor, and when the MDF fiber line reaches the position right below the CCD camera 21, transmitting an arrival signal to the industrial personal computer 4 by the photoelectric sensor 5, controlling the conveyor to stop running and sending a photographing signal of the CCD camera 21 to the industrial personal computer 4;
(2) the CCD camera 21 takes a picture and transmits the acquired image to the industrial personal computer 4 through the image acquisition card 23;
(3) the industrial personal computer 4 preprocesses the acquired image in a 5G edge calculation layer, uploads the image to the cloud for calculation, and finally learns a plurality of defect modes in a PasS layer by combining a surface defect model, a point cloud technology and a deep learning model of the MDF fiber line, so that the surface defect of the MDF fiber line can be accurately detected; when the industrial personal computer 4 detects that the MDF fiber line has defects, the industrial personal computer 4 drives the seven-degree-of-freedom heavy-load mechanical arm to grab the defective MDF fiber line to the unqualified product area.
Wherein, the pretreatment comprises the following steps:
(1) graying: converting the color image into a gray image by adopting a weighted average gray method for the acquired image;
(2) and (3) straight line detection: carrying out linear detection on the grayed image by an LSD linear detection method;
(3) and (3) inclination correction: screening the image after the straight line detection according to conditions, acquiring a long edge of an MDF fiber line in the image, calculating a difference value of coordinates at two ends of the long edge in the horizontal and vertical directions, and obtaining a slope k and an inclination angle theta of the straight line where the long edge is located, so as to perform image inclination correction;
(4) edge detection: removing noise of the image after inclination correction by adopting a Gaussian filter, convolving each pixel point in the image by a Gaussian kernel, replacing the value of a central pixel point with the weighted average gray value of the pixel in the region, respectively calculating the amplitude values of the pixel points in the x direction and the y direction by adopting a Sobel operator, convolving the image with the kernel of the size of 3 multiplied by 3, calculating the gradient in the horizontal direction and the vertical direction, and solving the approximate gradient and the gradient direction; a non-maximum value inhibition method is adopted for the gradient amplitude, a local maximum value is searched, and non-edge pixels are eliminated redundantly; comparing and judging the amplitude of each pixel point with a set hysteresis threshold, wherein the pixel point is an edge pixel when the amplitude of the pixel point is greater than a high threshold, and is excluded when the amplitude of the pixel point is less than a low threshold, and if the amplitude is between the high threshold and the low threshold, the pixel is reserved only when the pixel is adjacent to a pixel which is higher than the high threshold;
(5) template matching: and (3) learning various defect modes by combining the image subjected to edge detection with a surface defect model, a point cloud technology and a deep learning model of the MDF fiber line in the PasS layer, and judging the surface defects of the MDF fiber line.
After the industrial personal computer 4 acquires the image of the fiber line of the MDF to be measured, because in the image acquisition process, the phenomena of poor illumination, irregular operation and the like inevitably occur, which all cause the deviation of the acquired image from the ideal, such as noise, the need of correcting the position of the image and the like, and the problems can cause great interference on the accuracy of the image analysis. In order to better identify the characteristic information to be detected in the image of the MDF fiber line, the image needs to be preprocessed first, so that the useful information of the MDF fiber line is enhanced, and the subsequent defect identification and detection on the MDF fiber line are faster and more accurate.
The preprocessing steps are reasonably designed, and because the detection and identification of the surface defects of the MDF fiber lines are irrelevant to the color characteristics of the MDF fiber lines, the color images are converted into gray level images, so that the algorithm processing speed can be increased; in order to avoid inclination of the positions of the acquired MDF fiber lines of the pictures in different degrees and improve the accuracy of subsequent detection, the inclination correction of the images is realized according to the slope of a known horizontal straight line by searching for the known horizontal straight line; the edge features are used as important matching features, and the accuracy of matching between the picture and the surface defect model of the MDF fiber line can be improved through edge detection.
Further, the line detection includes the following steps:
(1) scaling the image to 80% of the original image through Gaussian down-sampling;
(2) calculating the gradient of each pixel point in the horizontal direction and the gradient of each pixel point in the vertical direction, solving a gradient angle according to the horizontal gradient and the vertical gradient, and calculating a gradient amplitude;
(3) setting the gradient value of each pixel point of the image between 0 and 1023, and creating 1024 linked lists; through gradient traversal of pixel points, pixel point coordinates with the same gradient value are placed into the same linked list, each linked list is sorted according to the sequence from large to small, and finally the head and the tail of each linked list are linked to form a large linked list;
(4) traversing the whole linked list, taking the coordinate position of a pixel point stored at the head of the linked list as a seed pixel, diffusing the seed pixel through a region growing algorithm according to the direction with similar gradient, and removing the diffused pixel point from the linked list;
(5) performing rectangle fitting on the pixel region obtained by diffusion;
(6) calculating the precision error of the fitted rectangle according to an NFA formula, and if the precision error meets the requirement, considering the rectangle as an effective straight line; and (5) repeating the step (4) until the traversal of the whole linked list is completed.
The specific working method of the invention is many, and the above description is only the preferred embodiment of the invention. It should be noted that the above examples are only for illustrating the present invention, and are not intended to limit the scope of the present invention. It will be apparent to those skilled in the art that various modifications can be made without departing from the principles of the invention and these modifications are to be considered within the scope of the invention.
Claims (6)
1. The device for detecting the surface defects of the MDF fiber lines is characterized by comprising a light source (1), a camera assembly (2), a camera bracket (3) and an industrial personal computer (4); the light source (1) comprises an annular LED positive light source (11) and a rectangular LED auxiliary light source (12), the camera component (2) comprises a CCD video camera (21), a lens (22) and an image acquisition card (23), and the camera support (3) is of a double-column gantry structure and is arranged above a conveyor for conveying MDF fiber lines to perform surface defect detection; the annular LED positive light source (11) is arranged right above the camera support (3), and the rectangular LED auxiliary light sources (12) are arranged in the upper middle of the cylindrical surface of each upright of the camera support (3) facing one side of the conveyor; the CCD camera (21) is arranged in the middle of the bottom surface of the cross beam of the camera bracket (3); the CCD camera (21) is connected with an image acquisition card (23), a lens (22) is installed at the bottom of the CCD camera (21), and the image acquisition card (23) transmits shot image data to the industrial personal computer (2) in real time through a network port; the industrial personal computer (4) preprocesses image data in the 5G edge calculation layer, uploads the image data to the cloud for calculation, and finally learns multiple defect modes in the PasS layer by combining a surface defect model, a point cloud technology and a deep learning model of the MDF fiber line, so that the surface defect of the MDF fiber line is accurately detected.
2. The device for detecting the surface defects of the MDF fiber lines is characterized in that a photoelectric sensor (5) is arranged on the lower portion of a cylindrical surface of each upright of the camera support (3) facing one side of the conveyor, and the photoelectric sensor (5) is connected with an industrial personal computer (4) in a serial port mode.
3. The device for detecting the surface defects of the MDF fiber lines is characterized in that the CCD camera (21) adopts a high-resolution industrial digital CCD camera, and the lens (22) adopts a double telecentric machine vision lens.
4. The working method of the device for detecting the surface defects of the MDF fiber strands according to any one of claims 1-3, wherein the working method comprises the following steps:
placing an MDF fiber line to be subjected to surface defect detection on a conveyor, starting the conveyor, and when the MDF fiber line reaches the position right below a CCD camera (21), transmitting an arrival signal to an industrial personal computer (4) by a photoelectric sensor (5), controlling the conveyor to stop running and sending a photographing signal of the CCD camera (21) to the industrial personal computer (4);
the CCD camera (21) takes a picture and transmits the collected image to the industrial personal computer (4) through the image collecting card (23);
the industrial personal computer (4) preprocesses the acquired image in a 5G edge calculation layer, uploads the image to the cloud for calculation, and finally learns a plurality of defect modes in the PasS layer by combining a surface defect model, a point cloud technology and a deep learning model of the MDF fiber line, so that the surface defect of the MDF fiber line can be accurately detected; when the industrial personal computer (4) detects that the MDF fiber line has defects, the industrial personal computer (4) drives the seven-degree-of-freedom heavy-load mechanical arm to grab the defective MDF fiber line to the unqualified product area.
5. The working method of the device for detecting the surface defects of the MDF fiber lines is characterized in that the pretreatment comprises the following steps:
graying: converting the color image into a gray image by adopting a weighted average gray method for the acquired image;
and (3) straight line detection: carrying out linear detection on the grayed image by an LSD linear detection method;
and (3) inclination correction: screening the image after the straight line detection according to conditions, acquiring a long edge of an MDF fiber line in the image, calculating a difference value of coordinates at two ends of the long edge in the horizontal and vertical directions, and obtaining a slope k and an inclination angle theta of the straight line where the long edge is located, so as to perform image inclination correction;
edge detection: removing noise of the image after inclination correction by adopting a Gaussian filter, convolving each pixel point in the image by a Gaussian kernel, replacing the value of a central pixel point with the weighted average gray value of the pixel in the region, respectively calculating the amplitude values of the pixel points in the x direction and the y direction by adopting a Sobel operator, convolving the image with the kernel of the size of 3 multiplied by 3, calculating the gradient in the horizontal direction and the vertical direction, and solving the approximate gradient and the gradient direction; a non-maximum value inhibition method is adopted for the gradient amplitude, a local maximum value is searched, and non-edge pixels are eliminated redundantly; comparing and judging the amplitude of each pixel point with a set hysteresis threshold, wherein the pixel point is an edge pixel when the amplitude of the pixel point is greater than a high threshold, and is excluded when the amplitude of the pixel point is less than a low threshold, and if the amplitude is between the high threshold and the low threshold, the pixel is reserved only when the pixel is adjacent to a pixel which is higher than the high threshold;
template matching: and (3) learning various defect modes by combining the image subjected to edge detection with a surface defect model, a point cloud technology and a deep learning model of the MDF fiber line in the PasS layer, and judging the surface defects of the MDF fiber line.
6. The working method of the device for detecting the surface defects of the MDF fiber lines according to claim 4, wherein the straight line detection comprises the following steps:
(1) scaling the image to 80% of the original image through Gaussian down-sampling;
(2) calculating the gradient of each pixel point in the horizontal direction and the gradient of each pixel point in the vertical direction, solving a gradient angle according to the horizontal gradient and the vertical gradient, and calculating a gradient amplitude;
(3) setting the gradient value of each pixel point of the image between 0 and 1023, and creating 1024 linked lists; through gradient traversal of pixel points, pixel point coordinates with the same gradient value are placed into the same linked list, each linked list is sorted according to the sequence from large to small, and finally the head and the tail of each linked list are linked to form a large linked list;
(4) traversing the whole linked list, taking the coordinate position of a pixel point stored at the head of the linked list as a seed pixel, diffusing the seed pixel through a region growing algorithm according to the direction with similar gradient, and removing the diffused pixel point from the linked list;
(5) performing rectangle fitting on the pixel region obtained by diffusion;
(6) calculating the precision error of the fitted rectangle according to an NFA formula, and if the precision error meets the requirement, considering the rectangle as an effective straight line; and (5) repeating the step (4) until the traversal of the whole linked list is completed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110731154.8A CN113588655A (en) | 2021-06-30 | 2021-06-30 | Detection device for surface defects of MDF fiber lines and working method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110731154.8A CN113588655A (en) | 2021-06-30 | 2021-06-30 | Detection device for surface defects of MDF fiber lines and working method thereof |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113588655A true CN113588655A (en) | 2021-11-02 |
Family
ID=78245079
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110731154.8A Withdrawn CN113588655A (en) | 2021-06-30 | 2021-06-30 | Detection device for surface defects of MDF fiber lines and working method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113588655A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117291918A (en) * | 2023-11-24 | 2023-12-26 | 吉林大学 | Automobile stamping part defect detection method based on three-dimensional point cloud |
-
2021
- 2021-06-30 CN CN202110731154.8A patent/CN113588655A/en not_active Withdrawn
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117291918A (en) * | 2023-11-24 | 2023-12-26 | 吉林大学 | Automobile stamping part defect detection method based on three-dimensional point cloud |
| CN117291918B (en) * | 2023-11-24 | 2024-02-06 | 吉林大学 | Automobile stamping part defect detection method based on three-dimensional point cloud |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109454006B (en) | Detection and classification method based on device for online detection and classification of chemical fiber spindle tripping defects | |
| CN110314854B (en) | Workpiece detecting and sorting device and method based on visual robot | |
| CN109550712B (en) | Chemical fiber filament tail fiber appearance defect detection system and method | |
| CN109255787B (en) | System and method for detecting scratch of silk ingot based on deep learning and image processing technology | |
| CN105403147B (en) | One kind being based on Embedded bottle embryo detection system and detection method | |
| CN110310255B (en) | Point switch notch detection method based on target detection and image processing | |
| CN104077577A (en) | Trademark detection method based on convolutional neural network | |
| CN101105459A (en) | Empty bottle mouth defect inspection method and device | |
| CN102175692A (en) | System and method for detecting defects of fabric gray cloth quickly | |
| CN104483320A (en) | Digitized defect detection device and detection method of industrial denitration catalyst | |
| CN107084992A (en) | A kind of capsule detection method and system based on machine vision | |
| CN106709529B (en) | Visual detection method for photovoltaic cell color difference classification | |
| CN110186375A (en) | Intelligent high-speed rail white body assemble welding feature detection device and detection method | |
| CN104132945A (en) | On-line surface quality visual inspection device for bar based on optical fiber conduction | |
| CN202351182U (en) | Online high-speed detection system for defects on surface of tinplate | |
| CN113588655A (en) | Detection device for surface defects of MDF fiber lines and working method thereof | |
| CN111330874A (en) | Detection device and detection method for pollution or impurity defects of bottom area of medicine bottle | |
| CN105973903B (en) | A kind of Oral liquid bottle lid detection method | |
| CN105699386B (en) | A kind of automatic cloth inspection labeling method using contact-type image sensor | |
| CN117214178A (en) | Intelligent identification method for appearance defects of package on packaging production line | |
| CN107121063A (en) | The method for detecting workpiece | |
| CN208155893U (en) | binocular vision scratch detection system based on high speed camera | |
| CN105738376B (en) | A kind of automatic cloth inspecting machine using contact-type image sensor | |
| CN115753765A (en) | Method for identifying position and posture of cheese | |
| CN115082504A (en) | Light spot identification method for solar photovoltaic panel |
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 | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20211102 |
|
| WW01 | Invention patent application withdrawn after publication |