Study Smooth Reflective Surfaces Automatic Defect Detection Method

With the development of manufacturing technology continuously improved, the requirements for sanitary products is also rising.The tap as the main structure of bathroom products, its quality will affect the appearance and life of bathroom products. Faucet as the main structure of bathroom products, its quality will affect the appearance and life of bathroom products. Therefore, defect detection must be do to the surface of the workpiece.At present, BM Sanitary Ware Technology Co.Ltd mainly depend on artificial detection. While the artificial detection is low work efficiency, the labor intensity of workers is high, the reliability of detection depends on many subjective factors. This will not only increase the management and labor cost, and it is difficult to guarantee the quality of faucet, high speed production is not suitable for the modern enterprise. It is not suitable for the high-speed production of modern enterprise. Therefore,it is urgent need to research for an automatic detect method for faucet surface defection. It will enable the faucet surface defects to be efficient, automatic and fast, then increase the competitiveness of enterprises.This paper intensive studied image processing and machine vision technology, with the faucet surface quality inspection for example,stressed sdudied on the image collection of smooth reflective surface and image defect recognition algorithm. Put forward a set of feasible solution, in the case of faucet, develop a feasible automatic detection system for smooth reflective surface defects. The main work is as follows:1.Obtained the defect standard and the main defect types of faucet by do field investigation to the production and processing process of faucet.2.Collect effectively original image. The surface of faucet is smooth and high reflective, so the external environment had great influence on image collection. Because the surface is smooth and it with surface structure, it’s easily bring diffuse reflection, it will cover the defect. The light deficiency of the edge will made the image be too dark, which may cover the defect. Therefore, a chamber environment and appropriate light source is used for the image acquisition system, then take photos from multiple angles. Using linear CCD and self-made video platform to improve the situation, to maintain the stability of the system.3.According to the analysis of the image features, some image processing is do to the original image. Which can remove noises and irrelevant background of the image,then enhanced the image.To the different characteristics of the defects in the different parts this paper choose effective method for image analysis and feature extraction.4.According to the type of defect, comprehensive use of various image processing and recognition technology, the paper propose the method of detect and identify the surface with smooth and reflective feature. And do simulation under MATLAB(R2008a).5.Using MATLAB programming realize the simulation of defect detection and identification system.The surface detection system described in this article used for the smooth and reflective object surface hast the characteristics of fast, standardized, efficient and intelligent. And the simulation results show that this system can obtain ideal effect.