Open AccessA surface defect identification method based on improved threshold segmentation algorithm
Author(s) -
Xinglong Feng,
Xianwen Gao,
Ling Luo
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1651/1/012072
Subject(s) - segmentation , simplicity , identification (biology) , computer science , quality (philosophy) , artificial intelligence , image (mathematics) , image segmentation , emphasis (telecommunications) , algorithm , pattern recognition (psychology) , data mining , botany , biology , telecommunications , philosophy , epistemology
For enterprises, defect detection is very important because it is related to the quality of products produced by enterprises. With the development of machine vision, accurate analysis of image data benefits defect detection. In an enterprise that produces electronic cigarettes, professional and technical personnel used to detect a defect in a workpiece by using manual testing. The defect detection rate of this method is only 95%, and the efficiency is low. We use an improved threshold segmentation method to solve this problem in this paper and we have achieved success. Compared with typical methods, the accuracy of our proposed algorithm reaches over 99% and at the same time, the detection efficiency has been improved by more than 50%. Our method also has the advantage of simplicity, practicality and low cost.