Assembly Line Error Proofing Detection Technology Research Based On Machine Vision

The study of error proof detection technology for automatic assembly line is of great significance for improving the quality of product and production efficiency.Machine vision,as a branch of artificial intelligence,is used to improve the flexibility of assembly lines and the level of industrial automation in the error detection of assembly lines.In this paper,a variety of different requirements for error proof detection in the assembly line of the gearbox and motor controller are introduced and summarized as color detection,presence detection,location detection and number detection.The hardware composition and image acquisition technology of the visual inspection system are analyzed,and a test platform is built on the actual assembly line.Using image samples of spare parts which are collected from assembly lines to research on the effect of image preprocessing methods and parameters to process result and to do application research on the extraction,recognition and matching algorithm of edge feature,histogram feature,and grayscale feature.A visual error proof detection software is designed by the use of modern software engineering design method and based on C# language and.NET platform as well as EmguCV vision tool library.For the feature s of different detection objects,using visual tools module software package,the images of assembly line parts are processed by image processing and feature recognition and detection respectively.Finally,through the practical application test and data analysis,it is verified that the visual error proofing software has the characteristics of convenient configuration,stable and reliable operation and simple operation.It has a certain reference value as the application of machine vision error proofing technology.