CN-116428981-B - Online plate detection method and system

Abstract

The invention discloses an online plate detection method and system, which belong to the technical field of plate detection and comprise a size detection step and/or a hole site detection step, wherein the size detection of a plate comprises the steps of carrying out region segmentation treatment and profile extraction treatment on a detection image to obtain an edge profile of the plate, carrying out coordinate conversion treatment on the edge profile to obtain physical size information of the plate, and the hole site detection of the plate comprises the steps of carrying out profile extraction treatment on a plate region, screening according to a preset profile target to determine the hole site profile of the plate and calculating the physical hole site information of the plate according to the hole site profile. According to the invention, the size detection processing and the hole position detection processing are respectively carried out on the plates with different materials, so that the plates with different colors and different roughness are considered, noise points of images with different materials are effectively filtered, the influence of different materials on algorithms such as image segmentation, contour extraction and the like is reduced, and the size and the hole position detection precision are ensured.

Inventors

• ZHANG SHIFU

Assignees

• 成都天衡智造科技有限公司

Dates

Publication Date

20260512

Application Date

20230417

Claims (7)

1. 1. The on-line plate detection method is characterized by comprising a size detection step and a hole position detection step, wherein the size detection comprises a size detection substep of a high-reflection plate and a size detection substep of a matte material plate; the size detection of the highly reflective sheet material comprises: Performing region segmentation treatment and contour extraction treatment on the detection image to obtain the edge contour of the plate; performing coordinate conversion processing on the outline of the edge to obtain physical size information of the plate; The size detection of the matt material plate comprises: Performing region segmentation treatment and contour extraction treatment on the detection image to obtain the edge contour of the plate; Smoothing the edge profile of the plate, and calculating the plate dimension length L1 and the plate dimension width W1 by adopting rectangular fitting and a least square method, wherein the plate dimension length L2 and the plate dimension width W2 are respectively adjusted in a self-adaptive manner by using rectangular fitting weights and least square method weight coefficients, so that the optimal profile is obtained by fitting; performing coordinate conversion processing on the optimal outline to obtain physical dimension information of the plate; hole site detection of high reflection of light panel includes: performing region segmentation processing on the detection image to determine a plate region; Carrying out contour extraction processing on the plate area, screening according to a preset contour target, and determining the hole site contour of the plate; calculating physical hole site information of the plate according to the hole site profile; Hole site detection of matte material panel includes: performing region segmentation processing on the detection image to determine a plate region; Carrying out contour extraction processing on the plate area, screening according to a preset contour target, and determining the hole site contour of the plate; Carrying out contour smoothing treatment on the hole site contour, adopting circle fitting and maximum inscribed circle fitting to respectively calculate the hole site dimension r1 and coordinates (x 1, y 1), the hole site dimension r2 and coordinates (x 2, y 2), and carrying out self-adaptive adjustment on the circle fitting weight and the maximum inscribed circle fitting weight coefficient to obtain the optimal hole site contour; calculating physical hole site information of the plate according to the optimal hole site profile; The region segmentation processing and the contour extraction processing of the detection image comprise the following steps: Carrying out threshold processing on the detected image, screening a plate area, and converting the plate area into a plate contour; sequentially performing expansion and contraction treatment on the contour of the plate, and further extracting an edge contour image of the plate; performing contour fitting treatment on the edge contour image to obtain the edge contour of the plate; The performing contour smoothing on the hole site contour comprises the following steps: Carrying out average gray value calculation on the hole site contour image to obtain an average gray value K1; Searching a binarization threshold k2 of the hole site contour image by adopting a histogram method; average value processing is carried out on the average gray value K1 and the binarization threshold value K2 to obtain a binarization processing threshold value; performing binarization processing on the hole site profile image to obtain an optimal hole site profile; the method further comprises the step of judging the plate quality: and calculating the average gray value of the plate area, wherein the plate area is a high-reflection plate if the average gray value is greater than or equal to a threshold value, and the plate area is a matte plate if the average gray value is less than the threshold value.
2. 2. The method for online board inspection as recited in claim 1, wherein the step of screening the board according to the predetermined profile target comprises: Screening the extracted contour according to a preset contour target to obtain a preliminary contour meeting the length characteristics; performing circularity screening on the preliminary contour to obtain a circular contour; And fitting the circular profile to obtain the hole site profile of the plate.
3. 3. The method for online board inspection as recited in claim 1, further comprising: Determining a label area of the target image; identifying the tag area to obtain standard size information and standard hole site information of the plate; and judging whether the physical size information and the physical hole site information of the plate are qualified or not according to the standard size information and the standard hole site information.
4. 4. The method for online sheet detection according to claim 3, wherein determining the tag area of the target image comprises the steps of: a blob region segmentation algorithm is adopted to obtain a black rectangular region of the high-reflection plate image, and then a tag region is determined; and acquiring a white rectangular area of the matte material plate image by adopting a blob area segmentation algorithm, and further determining a label area.
5. 5. The on-line plate detection system is characterized by comprising a central calculation control unit, wherein a size detection module and a hole position detection module are integrated on the central calculation control unit, the size detection module comprises a high-reflection plate size detection sub-module and a matte material plate size detection sub-module, and the hole position detection module comprises a high-reflection plate hole position detection sub-module and a matte material plate hole position detection sub-module; The high-reflection plate size detection submodule is used for carrying out region segmentation treatment and contour extraction treatment on the detection image to obtain the edge outline of the plate; The matte material plate size detection submodule is used for carrying out region segmentation treatment and contour extraction treatment on a detection image to obtain an edge outline of the plate, carrying out smoothing treatment on the edge outline of the plate, adopting rectangular fitting and a least square method to respectively calculate the plate size length L1 and the plate size width W1, carrying out self-adaptive adjustment on the plate size length L2 and the plate size width W2, carrying out rectangular fitting weight and a least square method weight coefficient, carrying out fitting to obtain an optimal outline, and carrying out coordinate conversion treatment on the optimal outline to obtain physical size information of the plate; the high-reflection plate hole site detection submodule is used for carrying out region segmentation processing on the detection image to determine a plate region, carrying out contour extraction processing on the plate region, screening according to a preset contour target to determine the hole site contour of the plate, and calculating physical hole site information of the plate according to the hole site contour; The matte material plate hole site detection submodule is used for carrying out region segmentation processing on a detection image to determine a plate region, carrying out contour extraction processing on the plate region, screening according to a preset contour target to determine the hole site contour of the plate, carrying out contour smoothing processing on the hole site contour, adopting circle fitting and maximum inscribed circle fitting to respectively calculate the hole site dimension r1 and coordinates (x 1, y 1), the hole site dimension r2 and coordinates (x 2, y 2), and carrying out self-adaptive adjustment on the circle fitting weight and the maximum inscribed circle fitting weight coefficient to obtain the optimal hole site contour; The region segmentation processing and the contour extraction processing of the detection image comprise the following steps: Carrying out threshold processing on the detected image, screening a plate area, and converting the plate area into a plate contour; sequentially performing expansion and contraction treatment on the contour of the plate, and further extracting an edge contour image of the plate; performing contour fitting treatment on the edge contour image to obtain the edge contour of the plate; The performing contour smoothing on the hole site contour comprises the following steps: Carrying out average gray value calculation on the hole site contour image to obtain an average gray value K1; Searching a binarization threshold k2 of the hole site contour image by adopting a histogram method; average value processing is carried out on the average gray value K1 and the binarization threshold value K2 to obtain a binarization processing threshold value; performing binarization processing on the hole site profile image to obtain an optimal hole site profile; The central computing control unit is also used for judging the plate material quality: and calculating the average gray value of the plate area, wherein the plate area is a high-reflection plate if the average gray value is greater than or equal to a threshold value, and the plate area is a matte plate if the average gray value is less than the threshold value.
6. 6. The on-line board inspection system of claim 5, further comprising: The positioning detection unit is connected with the central calculation control unit and used for detecting the position information of the plate on the conveyor belt; The image acquisition unit is connected with the central calculation control unit and is used for acquiring image information of the plate on the conveyor belt.
7. 7. The on-line board inspection system of claim 6, wherein the image acquisition unit comprises a first image acquisition module and a second image acquisition module; The first image acquisition module is used for acquiring target images required by tag identification and comprises a linear array camera and a highlight light source, wherein the linear array camera is vertically arranged with the conveyor belt, and the highlight light source is arranged at an angle of 35-50 degrees with the conveyor belt; the second image acquisition module is used for acquiring detection images required by size detection and hole site detection and comprises a wide-width linear array camera and a customized light source, wherein the wide-width linear array camera is vertically arranged with the conveyor belt, and the customized light source is arranged at an angle of 85-95 degrees with the conveyor belt.

Description

Online plate detection method and system Technical Field The invention relates to the technical field of plate detection, in particular to an online plate detection method and system. Background In the production process of the customized furniture, the dimensional accuracy of the product cannot be achieved, and the installation of the furniture and the attractive requirements on the appearance are seriously influenced. The most important step for determining the accuracy of the plate is the cutting correction and edge sealing process of the formed plate. Therefore, the detection of the appearance size of the plates is of great significance in ensuring the quality of each plate. In the traditional furniture industry, the dimension detection after edge sealing and molding of the plate is also dependent on manual sampling inspection. However, the manual spot check mode is difficult to ensure accurate measurement of all products, and for large-size plates, the detection accuracy can not reach 0.5mm, the efficiency is low, the detection time is long, the accuracy of a detection tool (tape) can not be ensured, and the like, and the requirements of an automatic production line can not be met. The detection method of machine vision can solve the problems existing in the prior factories. Compared with the manual detection method, the machine vision detection has the advantages of no contact, high precision, high speed and the like. The measured plate sizes of the plate detection system are matrix plates with different specifications, the produced materials of different plates can change, the surface color of the plates can also correspondingly change, and the surface textures of the plates also have differences. Meanwhile, the plate surface and the edge banding are different in material, and the edge banding and the plate surface are subjected to arc chamfering, so that the detection precision of the existing machine vision-based detection algorithm is greatly reduced. Therefore, how to improve the detection accuracy of the detection algorithm to meet the requirements of the client is a technical problem that needs to be solved at present. Disclosure of Invention The invention aims to overcome the problems in the prior art and provides an online plate detection method and system. The invention aims at realizing the technical scheme that the method comprises a dimension detection step and/or a hole position detection step, wherein the dimension detection comprises a dimension detection substep of a high-reflection plate and/or a dimension detection substep of a matte material plate; The dimension detection of the high-reflectivity plate comprises the steps of carrying out region segmentation treatment and contour extraction treatment on a detection image to obtain the outline of the edge of the plate; The size detection of the matte material plate comprises the steps of carrying out region segmentation treatment and contour extraction treatment on a detection image to obtain the edge outline of the plate, carrying out smoothing treatment on the edge outline of the plate, fitting to obtain the optimal outline, and carrying out coordinate conversion treatment on the optimal outline to obtain the physical size information of the plate; the hole site detection of the high-reflectivity plate comprises the steps of carrying out region segmentation processing on a detection image to determine a plate region, carrying out contour extraction processing on the plate region, screening according to a preset contour target to determine the hole site contour of the plate, and calculating physical hole site information of the plate according to the hole site contour; The hole site detection of the matte material plate comprises the steps of carrying out region segmentation processing on a detection image to determine a plate region, carrying out contour extraction processing on the plate region, screening according to a preset contour target to determine the hole site contour of the plate, carrying out contour smoothing processing on the hole site contour to obtain an optimal hole site contour, and calculating physical hole site information of the plate according to the optimal hole site contour. In an example, the performing the region segmentation process and the contour extraction process on the detected image includes: Carrying out threshold processing on the detected image, screening a plate area, and converting the plate area into a plate contour; sequentially performing expansion and contraction treatment on the contour of the plate, and further extracting an edge contour image of the plate; and performing contour fitting treatment on the edge contour image to obtain the edge contour of the plate. In an example, the filtering according to the preset profile objective includes: Screening the extracted contour according to a preset contour target to obtain a preliminary contour meeting the length characteristics; performing c