A Target Defect Extraction Method

Abstract

The present invention provides a target defect extraction method, comprising: step 1) selecting data from the actual gray histogram of the target image for fitting to obtain the standard gray histogram of the target image; step 2) obtaining the actual gray histogram The first intersection point and the last intersection point of the grayscale histogram and the standard grayscale histogram; Step 3) determine two grayscale values ​​as thresholds according to the first intersection point and the last intersection point; The target image is subjected to threshold segmentation to obtain suspected defects. The invention can extract suspected defects more completely, has higher precision and reduces the phenomenon of over-segmentation.

Description

technical field [0001] The present invention relates to the technical field of image processing, and more specifically, to a method for extracting suspected defects of a target from an image containing the target. Background technique [0002] With the development of intelligence, industrialization and mechanization, bearings have become an indispensable part of mechanical equipment. Since bearings are used in every corner of life, once a bearing has a problem, it will bury hidden dangers for mechanical failure, and even cause mechanical failure and cause time and economic losses. More seriously, it may lead to a series of accidents. Therefore, how to ensure the quality of bearings is particularly important. In order to ensure the quality of bearings, the current commonly used technology is to perform defect detection on the captured bearing images to eliminate and replace problematic bearings. [0003] At present, the defect detection of the target in the image usually in...

AI Technical Summary

Problems solved by technology

If the suspected defects obtained by defect extraction are incomplete, it will lead to missed detection and affect the overall result of defect detection

Commonly used defect extraction methods include threshold segmentation, which can be further divided into single-threshold segmentation and multi-threshold segmentation. Among them, the former (including OTSU segmentation method, maximum entropy segmentation method, etc.) Suspected defects with high gray levels may lead to the problem of missing defects; while multi-threshold segmentation with more than two thresholds will lead to complex calculations and low real-time performance, and too many thresholds may also cause the extracted The defect information is incomplete

Images