Levelling valve has the characteristics of many types of products and high similarity.The traditional manual visual inspection method is difficult to classify its models reliably and effectively,resulting in high rework rate and low production efficiency.Therefore,in this paper,we proposed a method of levelling valve model identification by machine vision.First,using Python-OpenCV to design the method of image pre-processing algorithm and feature extraction.Then,using feature information processing and analysis to establish the model classification recognition algorithm and using this algorithm to develop the levelling valve recognition system.Finally,an example is made to verify the feasibility and effectiveness of the proposed method.
维普
万方数据