Study On Impurity Detection In Limpidity Liquid Base On Image Process

Product quality issue has been the focus of attention, and quality is the survival of the enterprise, the development of our brand of the country, and therefore, product quality inspection process is indispensable. The current prevalence of bottled liquid impurity detection method mainly the seizure of artificial lights, light scattering, high performance liquid chromatography, based on image processing methods. Based on image processing and the detection method is divided into two categories: single image-based methods and sequence (video) image-based method. In recent years, with the development of image processing technology, based on image processing technology of the bottled liquid impurity detection method in the detect effect and other costs relative to other aspects of the method has great advantages. Therefore, we designed an impurity detection system based on image processing , to detect impurity in limpidity Liquid of the various methods of in-depth study.First, according to the bottled liquid impurity detection character and the actual production requirements, design a set based on image processing of bottled liquid impurity detection system. Describe the principle of the whole system, workflow and the various components of the functions specifically.Secondly, access to the images and character of impurity carried out a detailed analysis; Introduced Several edge detection algorithm, According to the actual situation we use Sobel operator to detect the edge of the liquid; Introduction the Hough transform technology, and on the based on this presents a window on default and Hough transform the liquid regional segmentation method.Finally, compare and analysis the impurity detection method based on single image and sequence (video) image. Based on the single image impurity detection, analysis of gray image threshold segmentation and color images threshold segmentation, finally Come to a conclusion: The method based on single image impurity detection is only for special containers, does not apply to ordinary liquid impurity detection; For the impurity detection based on sequence (video) images, analysis the existing optical flow method, the elimination of background method, and classical frame difference method, and Raised an improved frame difference method to solve the problem of the noise-sensitive,easily lead to image hole phenomenon,the test results more than the actual number of impurity, and other issues of classic frame difference method.