Bottleneck defect detection method adopting residual analysis and dynamic threshold segmentation

Abstract

The invention discloses a bottleneck defect detection method adopting residual analysis and dynamic threshold segmentation. The method comprises the following steps: positioning a bottleneck area by using a randomized circle assessment technology; carrying out strong smoothing treatment on a bottleneck target image obtained through polar coordinate transformation expansion, differentiating bottleneck target images before and after smoothing treatment to form a threshold curved surface changing with the gray value of the original bottleneck target image, and carrying out dynamic threshold segmentation on the bottleneck target image of the curved surface; and carrying out area connectivity detection on a binary image obtained after the segmentation, and judging whether defects exist or not according to the height, the width and the area of the connection area. The method has very strong adaption ability to the change and interference of the gray value of an identification target in the image, has a fast execution speed, and effectively solves the problem of high-speed and high-precision detection of the bottleneck defects.

Description

technical field [0001] The invention belongs to the field of image processing, and relates to a bottleneck detection method for residual analysis and dynamic threshold segmentation. Background technique [0002] my country's annual demand for wine bottles is huge. According to industry data, the cumulative output of my country's beer industry in 2014 was as high as 49.2185 million kiloliters. Calculated according to the bottle capacity of 530ml per bottle of beer, it would require as much as 9.28651x10 10 However, more than 80% of the beer bottles use recycled old bottles, and there are a large number of bottles with damaged bottle mouths. The use of bottles with damaged bottle mouths may bring major safety hazards to the production line and consumers. For defect detection, it is a necessary process to eliminate bottles with unqualified bottle necks. At present, there are a large number of methods for bottle neck defect detection at home and abroad. [0003] In foreign coun...

AI Technical Summary

Problems solved by technology

[0005] The present invention aims at the problem of low detection accuracy of bottle mouth defects in the above prior art, and proposes a bottle mouth defect detection method based on residual analysis and dynamic threshold segmentation

Images