Search

CN-122023320-A - Intelligent detection method and system for defects of paper product printed patterns

CN122023320ACN 122023320 ACN122023320 ACN 122023320ACN-122023320-A

Abstract

The invention relates to the technical field of printing detection, and discloses an intelligent detection method and system for pattern defects of paper products, which are characterized in that pattern image data of the paper products after printing are collected and subjected to grey treatment, geometric correction and illumination balance treatment are carried out on the grey treated image, the size and the position of the illumination balance treated image are aligned with those of standard grey images in an image library, error degree WC is calculated, when the calculated error degree WC is larger than a preset error threshold value Y, an alarm is triggered and the defect position is positioned, according to the invention, after image data acquisition, gray processing, geometric correction and illumination balancing processing are respectively carried out, and after the processing, image quality interference caused by factors such as shooting angle deviation, paper physical deformation, site illumination non-uniformity and the like can be eliminated, so that the image to be detected and a standard image are ensured to have high comparability in space and illumination conditions, errors caused by environmental factors are reduced, and the defect detection precision is improved.

Inventors

  • Teng Xinxu
  • WANG MINGLIANG

Assignees

  • 徐州太平洋印务有限公司

Dates

Publication Date
20260512
Application Date
20260128

Claims (10)

  1. 1. The intelligent detection method for the defects of the printing patterns of the paper products is characterized by comprising the following steps of: S1, collecting pattern image data of a printed paper product, and carrying out graying treatment on the collected data; s2, carrying out geometric correction on the image after graying based on the standard image and the real-time image; S3, performing illumination equalization processing on the corrected image limited contrast self-adaptive histogram equalization; s4, aligning the size and the position of the image subjected to the illumination balancing treatment with the standard gray level image in the image library; S5, calculating error degree WC between the aligned contrast image and the standard image; and S6, triggering an alarm and positioning the defect position when the calculated error degree WC is larger than a preset error threshold Y.
  2. 2. The intelligent detection system for the defects of the paper product printing patterns is characterized by comprising an image library, an acquisition unit, a gray level unit, a correction unit, an equalization unit, an alignment unit, a calculation unit and an alarm unit, wherein the image library is used for storing standard images of patterns, the acquisition unit is used for acquiring image data information of the printed patterns, the gray level unit is used for carrying out gray level processing on the acquired image data, the correction unit is used for carrying out geometric correction on the image subjected to gray level processing, the equalization unit is used for carrying out illumination equalization processing on the image subjected to gray level processing, the alignment unit is used for aligning the standard images in the image library with the comparison image subjected to the equalization unit, the calculation unit is used for calculating error degree WC, and the alarm unit is used for carrying out alarm and alarm position positioning when the error degree WC is greater than an error threshold value Y.
  3. 3. The intelligent detection system for defects of paper product printed patterns according to claim 2, wherein the gray unit performs gray processing on the image data collected by the collecting unit, a gray processing formula is hd=0.3 r+0.58g+0.12b, HD is a gray value after processing, R is a red image, G is a green image, B is a blue image, R, G, B are all in a gray value range of 0-255, the gray unit sends the image data after gray processing to the correcting unit, and the gray processing unit performs gray processing on the patterns in an image library and stores the patterns in the image library.
  4. 4. The intelligent detection system for defects in paper product printed patterns according to claim 2, wherein the correction unit receives the image data after graying and performs geometric correction, the correction unit retrieves an initial reference feature set of a current pattern from an image library, extracts the current reference feature set of the image data after graying, calculates an affine transformation matrix mapping a real-time image space to a standard image space based on the initial reference feature set and the current reference feature set, performs geometric transformation on the image data after graying by adopting an inverse matrix of the affine transformation matrix, calculates a corresponding source coordinate of the image data after graying in an original real-time image by adopting an inverse matrix for each target pixel position in the transformed image, acquires a pixel gray value of the source coordinate by adopting a bilinear interpolation method, and generates an aligned image after geometric correction.
  5. 5. The intelligent detection system for defects of paper product printed patterns according to claim 2, wherein the equalization unit receives the aligned image processed by the correction unit, the equalization unit performs image illumination equalization processing in a manner of limiting contrast adaptive histogram equalization, the equalization unit divides the geometrically corrected image into 8x8 rectangular areas, calculates a gray histogram of each small area separately, equalizes the cut histogram, distributes gray values of pixels in the rectangular areas to enhance local contrast, and obtains new gray values of each pixel point by bilinear interpolation from a transformation function of four sub-blocks around the pixel point.
  6. 6. The intelligent detecting system for paper product printing pattern defects according to claim 5, wherein each gray level histogram in the equalizing unit is provided with a contrast limiting threshold, the number of pixels with gray level values calculated by the histogram exceeding the contrast limiting threshold is uniformly distributed on all gray levels, and the equalizing unit sends the image after the illumination equalizing process to the aligning unit.
  7. 7. The intelligent detection system for defects in printed patterns of paper products according to claim 2 wherein the alignment unit receives the image processed by the equalization unit and retrieves a grayscaled image of the image from the image library, the alignment unit marks the image transmitted by the equalization unit as a contrast image and the grayscaled image in the image library as a standard image, the alignment unit aligns the standard image and the contrast image in the same size, and the alignment unit transmits the two images after alignment to the calculation unit.
  8. 8. The intelligent detection system for printing pattern defects on paper products according to claim 7, wherein said calculation unit calculates an error degree WC of the comparison image and the standard image, and the error degree WC has a calculation formula of In the formula, JF is mean square error, BZ1 is standard mean square error, JW is absolute error sum, BZ2 is standard absolute error sum, the calculating unit compares the error threshold Y in the expected interior of the calculated error degree WC, when the error degree WC is greater than the error threshold Y, the calculating unit sends an alarm instruction to the alarm unit, when the error degree WC is less than or equal to the error threshold Y, the calculating unit does not send an instruction, and the alarm unit receives the alarm instruction to send an alarm and positions the image position sent by the alarm instruction.
  9. 9. The intelligent detection system for paper product printing pattern defects as set forth in claim 8, wherein the calculation formula of the mean square error JF when calculating the error degree WC in said calculation unit is as follows Where H is the height of the image, W is the width of the image, C is the number of channels of the image, i is the row index of the image pixels, j is the column index of the image pixels, k is the channel index, Is in position for standard image At the pixel value of the kth channel, In position for contrast image At the pixel value of the kth channel.
  10. 10. The intelligent detection system for printing pattern defects on paper products according to claim 8, wherein the calculation formula of the absolute error sum JW is when calculating the error degree WC in the calculation unit In the following Is in position for standard image The pixel value at which it is located, In position for contrast image Pixel values at.

Description

Intelligent detection method and system for defects of paper product printed patterns Technical Field The invention relates to the technical field of printing detection, in particular to an intelligent detection method and system for defects of printing patterns of paper products. Background The quality of the printed pattern of the paper product directly affects the quality of the appearance and brand image of the product. In the printing production process, due to the reasons of mechanical errors, uneven printing ink, inaccurate overprinting, adhesion of sundries, deformation of materials and the like, various defects such as color deviation, pattern dislocation, wiredrawing, printing leakage, dirty points and the like are extremely easy to generate, and currently, the detection of the printing patterns in the industry mainly depends on manual visual inspection or a simple photoelectric sensor; With the development of machine vision technology, intelligent detection based on images has become an important direction of industrial quality inspection, and some existing detection schemes still have the following problems that after image data acquisition, the image data cannot be well preprocessed, so that the problem that the accuracy of subsequent comparison is affected due to interference caused by uneven visual angles, distances and illumination in the image acquisition process cannot be solved, when characteristic alignment in patterns is carried out, accurate matching of pixel levels is difficult to realize, when defects are judged, the judgment criterion often depends on single pixel difference, is insensitive to local fine defects or defects of specific types, lacks a comprehensive quantitative evaluation value, and is not accurate enough. Disclosure of Invention In order to overcome the above defects in the prior art, the embodiment of the invention provides an intelligent detection method and system for defects of paper product printing patterns, so as to solve the technical problems in the background art. In order to achieve the purpose, the invention provides the following technical scheme that the intelligent detection method for the defects of the printing patterns of the paper products is characterized by comprising the following steps: S1, collecting pattern image data of a printed paper product, and carrying out graying treatment on the collected data; s2, carrying out geometric correction on the image after graying based on the standard image and the real-time image; S3, performing illumination equalization processing on the corrected image limited contrast self-adaptive histogram equalization; s4, aligning the size and the position of the image subjected to the illumination balancing treatment with the standard gray level image in the image library; S5, calculating error degree WC between the aligned contrast image and the standard image; and S6, triggering an alarm and positioning the defect position when the calculated error degree WC is larger than a preset error threshold Y. The intelligent detection system for the defects of the printing patterns of the paper products is characterized by comprising an image library, an acquisition unit, a gray level unit, a correction unit, an equalization unit, an alignment unit, a calculation unit and an alarm unit, wherein the image library is used for storing standard images of the patterns, the acquisition unit is used for acquiring image data information of the printed patterns, the gray level unit carries out gray level processing on the acquired image data, the correction unit carries out geometric correction on the gray level processed image, the equalization unit carries out illumination equalization processing on the image after the correction, the alignment unit aligns the standard images in the image library with the comparison image after the equalization unit, the calculation unit is used for calculating error degree WC, and the alarm unit carries out alarm and alarm position positioning on the error degree WC > error threshold Y. In a preferred embodiment, the gray unit performs gray processing on the image data collected by the collecting unit, a gray processing formula is hd=0.3 r+0.58g+0.12b, where HD is a gray value after processing, R is a red image, G is a green image, B is a blue image, and R, G, B are all in a gray value range of 0-255, the gray unit sends the image data after gray processing to the correcting unit, and the gray unit stores the image after gray processing on the pattern in the image library. In a preferred embodiment, the correction unit receives the image data after the graying and performs geometric correction, the correction unit invokes an initial reference feature set of a current pattern from an image library, extracts the current reference feature set of the image data after the graying, calculates an affine transformation matrix for mapping the real-time image space to the standard image space based on t