A Fast Gap Detection Algorithm Based on Machine Vision

• Published in: Journal of physics. Conference series Vol. 2670; no. 1; pp. 12022 - 12027
• Main Authors: Li, Pengxin, Gao, Quanqin, Chen, Yuepeng, Sun, Haiyang, Cai, Qingyu, Tang, Xiaoan
• Format: Journal Article
• Language: English
• Published: Bristol IOP Publishing 01.12.2023
• Subjects: Algorithms; Extreme values; Hough transformation; Machine vision; Physics; Vision systems
• ISSN: 1742-6588; 1742-6596; 1742-6596
• DOI: 10.1088/1742-6596/2670/1/012022

In order to be able to quickly locate the position of the gap of an industrial device in the industrial inspection process, a fast gap detection algorithm based on machine vision is proposed by using the feature of a high gray level difference between the image gap and the background in machine vision. First, nonlinear equalization preprocessing is adopted, and then multiple one-dimensional functions are composed based on the sum of the gray values of the horizontal axis as well as the differentiation, and then the proximity position averaging operation is done after obtaining the positional information through the extreme value and the most value operation. The algorithm transforms the two-dimensional image into a combination of multiple one-dimensional functions with related features, and finally accurate position information is obtained by analysing and fusing the one-dimensional functions. Multiple discriminative bases are used for the same data source to ensure the accuracy of the results. Experiments show that the method proposed is faster and less error-prone, has higher accuracy and detection efficiency, and meets the needs of industrial real-time detection, compared with the traditional Hough transform algorithm and the LSD detection algorithm.