基于机器视觉的高度阀型号识别方法研究

Model Recognition&Assembly Error Proofing Method of Levelling Valve Based on Machine Vision

轨道交通用高度阀具有产品种类多、结构相似度极高等特点,传统人工目测方法难以对其型号进行可靠有效地识别分类,导致返工、发错货物等问题。因此,提出一种基于机器视觉的高度阀型号识别方法:首先,基于Python-OpenCV针对高度阀部件设计图像预处理算法和特征提取方法;其次,进行特征信息处理分析,建立型号分类识别算法;同时,基于以上算法研发了高度阀类型识别系统;最后,通过实例验证了本研究所提出方法的可行性和有效性。

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.