Artificial board surface defect detection method based on singular value decomposition

Abstract

The invention discloses an artificial board surface defect detection method based on singular value decomposition. The method comprises the following steps: firstly, conducting inter-row difference elimination processing on an original gray image of an artificial board; segmenting the grayscale image with the inter-row difference eliminated into n * n image blocks, carrying out singular value decomposition on each image block, taking a first singular value of the diagonal matrix as a feature value of the selected image block, storing the feature value into a feature matrix F, and carrying out statistics on a variance S and a mean value G of the feature matrix F to obtain a segmentation threshold T of the feature matrix F, T = S + lambda G; establishing a potential defect matrix, comparing each element value in the feature matrix F with a segmentation threshold T, if the element value is greater than the segmentation threshold T, setting the element value at the position corresponding to the potential defect matrix as 1, otherwise, setting the element value as 0; carrying out isolated point correction on the potential defect matrix; and if the number of the corrected element values in the defect matrix is 0, judging that the artificial board is qualified, otherwise, judging that the artificial board is unqualified.

Description

technical field [0001] The invention relates to an artificial board surface defect detection method using singular value decomposition, which belongs to the application field of digital image processing. Background technique [0002] With the continuous maturity of wood-based panel continuous press production line technology, the level of automatic production of wood-based panels has been continuously improved, but the detection of surface defects at the end is still dominated by manual visual recognition. There are many deficiencies in the way of manual naked eye detection, such as low recognition rate, low detection efficiency, and susceptibility to external factors, etc., and long hours of work are not conducive to the health of workers. The development of an automatic defect identification system has become an urgent need for the industry. The detection method using machine vision is a feasible method to automatically detect surface defects of wood-based panels. [000...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to be able to quickly and efficiently detect surface defects of wood-based panels, thereby providing a quality control method for wood-based panels that can adapt to on-line detection, so as to solve the high false detection rate and missed detection rate of traditional manual visual detection of wood-based panel surface defects, It is easy to cause visual fatigue of workers and other deficiencies

Images