Gear surface defect detection method and system based on image processing

Abstract

The invention relates to the technical field of image processing and mechanical part detection, in particular to a gear surface defect detection method and system based on image processing. The method comprises the following steps: extracting a single-tooth image from an initial gear image; classifying the pixel points in the single-tooth image according to the pixel values and the corresponding number of the pixel points to obtain boundary pixel points and surface pixel points; fitting the boundary pixel points to obtain a boundary curve; obtaining a boundary reference straight line according to the boundary pixel points on the boundary curve; determining whether the gear has a defect according to the similarity of the boundary reference straight line and the boundary curve; if the defect exists, performing up-down search by taking the boundary curve as a center to obtain boundary area edge points; determining a defect evaluation area through the boundary area edge points; and obtaining the gear wear degree according to the defect evaluation area. According to the method, the gear wear degree is evaluated by obtaining the boundary information, and the defect detection accuracy is ensured.

Description

Technical field [0001] The present invention relates to an image processing means for detecting mechanical and technical field, particularly to a method and system for defect detection image processing based on the gear surface. Background technique [0002] Mechanical gear affected by job content in the course of their work leading to regular wear and tear. Because of the large number of teeth of gear, so gear for wear defects requires a comprehensive analysis of every part of the gear. [0003] In the prior art, the optical device by the defect detection gear, but the high cost of this method, the light source in the actual scene may not meet the requirements, the impact detection efficiency. Using image processing and machine learning by the neural network of the defective pixel extracted, but the neural network requires a lot of tagging data for some of the annotation data difference is too large defects, defective pixel accuracy is not high, and can not accurately obtain the...

AI Technical Summary

Problems solved by technology

[0003] In the existing technology, the defects of gears can be detected by optical devices, but this method is costly, and the light source in the actual scene may not meet the requirements, which affects the detection efficiency

Image processing and machine learning methods can be used to extract defective pixels through the neural network, but the neural network requires a large amount of data labeling, and for some defects that are too different from the labeled data, the accuracy of the defective pixels is not high, and cannot Accurately obtain gear wear degree

Images