Automatic Monitoring Of Surface Defects Based On Image Recognition Of Automotive Filter Paper

With the rapid development of China’s automobile industry, the production and sales of car air cleaners are also greatly increased, in order to put an end to the fault of the traditional artificial detection filter, and greatly improve the detection efficiency, the automatic detection system of the automobile air filter paper surface defects was researched and improved. The architecture of the system is C/S, because of VC++ is object-oriented programming language for the development of image processing software provides a rich set of components, and is ideal for distributed development, so the future development tool selection VC.NET technology. And System development environment is Visual Studio.NET2005, background database is SQL Server 2010. This paper adopts the digital image recognition technology, the automation of filter paper detection achievement and high efficiency of the detection are through the process and recognition of the digital image, the measurement of identified filter paper’s defects, and the storage the information. After the image processing of preprocess, image enhancement, image segmentation, edge detection, erosion and image measurement, filter paper surface defect image will obtained the Binarized image. After measured the data of perimeter and area of the Binarized image of the defect store the defect image and calculation results to the database and provide control data to the positioning identify bodies. The filter paper defects was mainly detected in this paper, and it has the advantage of the characteristics of low investment costs, testing effect obvious, easy maintenance, low operating cost, so it can economically and effectively detect the paper defects.