The Intelligence Identify Method Of Glass Defect

With the social progress and development, glass products plays an indispensable effect in more and more industry, as an ever-increasing demand for materials in the production process of their quality control is particularly important, as it not only can save costs, increase market competitiveness and better reflect the state of the national high-tech development。The use of high-speed image acquisition technology, high-speed, high-precision defect detection algorithms and a combination of smart identification technology based on machine vision quality float glass line detection system, bringing enterprise is not just a technical change in the absolute。The so-called smart identification technology, have developed rapidly in today's scientific research areas, it is primarily the use of image feature extraction and classification combination of pairs of float glass extraction and classification of defects。Here we will study the following points to discuss the defects in float glass line detection system:(1), Image pre-processing, since the image acquisition process there will be noise, light, sort of external influence, will be to conduct a preliminary pre-processing to eliminate these effects.(2), Study of wavelet packets, wavelet, wavelet moment and many other feature extraction algorithms, to identify suitable feature extraction method of the glass.(3), Through the use of support vector machine classification and recognition methods, validation wavelet packet, wavelet and wavelet moments in the effect on the feature extraction.(4), On-line float glass defect inspection system of the research, design system solutions, will apply above algorithms to the system, so that it can achieve.