Answers are grounded in this paper and cite their sources.
An analysis related to the defect detection on a rubber weather-strip extrusion production line is presented herein. A vision system has been developed to evaluate the products quality in terms of dimensional accuracy. As a first method, the extrudate’s dimensions were identified by making use of image processing algorithms for detecting the edge and classifying the results. Furthermore, neural networks were used for identifying the dimensional patterns and classifying the profiles. Both methods have been proved to be accurate and repeatable. The study’s results show high potential for application in manufacturing environments due to models the high processing speed capabilities.
No takes yet. Share an insight, caveat, or question.
Stavropoulos et al. (2020) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: