Self-adaptive segmentation method based on crack image subarea gray scale mean value

Abstract

Disclosed is a self-adaptive segmentation method based on a crack image subarea gray scale mean value. The method comprises the following steps: 1), analyzing crack image characteristics on the basis of crack image gray scale statistical data of an image histogram; 2), selecting a narrow and long rectangular area as an image analysis ROI area, wherein the narrow and long rectangular area is symmetrical about a prefabricated crack expansion horizontal line and comprises fatigue cracks; 3), dividing the ROI area into two rows of multiple same-dimensioned square area image subareas from a prefabricated crack starting point to a crack right edge position; 4), calculating a grayscale distribution variance for each image subarea, and according to a variance distribution rule, excluding image subareas not including the fatigue cracks; 5), performing histogram calculation on the image subareas comprising the cracks to obtain a histogram two-dimensional array; and 6), performing crack segmentation on each subarea by use of a self-adaptive iteration method based on the crack image subarea gray scale mean value. According to the invention, the crack measuring precision is improved, the application scope is wide, the algorithm complexity is low, the realization is simple, and the operation speed is fast.

Description

technical field [0001] The invention relates to the field of image segmentation of image processing technology, in particular to an adaptive segmentation method based on the mean value of the gray level of a crack image sub-area to realize crack edge detection for subsequent crack length measurement. Background technique [0002] As we all know, machine vision technology based on image processing is applied to the measurement field of industrial products. While ensuring the measurement requirements, it makes the system non-contact, adaptable, fast, efficient, accurate, flexible, high reliability, easy operation and low cost. Low-level characteristics, the crack detection method using image processing technology can complete the crack detection and further analysis and processing intuitively, accurately, in real time and reliably. [0003] In order to ensure the accuracy of fatigue crack size measurement, the position of the crack edge must be strictly guaranteed in the image...

AI Technical Summary

Problems solved by technology

[0004] In order to overcome the disadvantages of poor reliability of existing crack edge detection methods, the present invention provides an adaptive segmentation based on crack image sub-area gray mean value with wide application range, low algorithm complexity, simple implementation and fast operation speed method, does not use any filter denoising and edge enhancement processing methods, but directly performs image segmentation and recognition on the original image

Images