Research On The Automatic Separation System Of Coal And Gangue Based On X-ray Image Processing

With the rapid development of science and technology, there is a growing emphasis on the use of clean coal, so the separation requirements of coal and gangue are getting higher. The sorting rates of traditional method of selection of artificial coal and wet are low and easily affected by people's health and other aspects. Wastewater generated by the wet-elect also caused some environmental pollution. In the emerging dry-elect methods, most of the recognition are based on the different gray of coal and gangue image under natural light and integration of its various features, which are easily affected by light environment, thereby making theirs sorting rates become unstable and having high computational complexity and sorting slow.X-ray as a relatively stable and insensitive to light environmental impact of imaging modalities is used in the separation system of coal and gangue, which overcomes the ever sensitive to natural light and forms the x-ray images of higher-quality. These images almost contain no noisy and are more easily and quickly processed. This paper is mainly based on x-ray image processing and designs feasible recognition algorithm which combined with the gray feature of images and the height of the object. The algorithm finally realizes the automatic separation of coal and gangue.The automatic recognition algorithm of coal and gangue based on X-ray image processing are mainly studied in this paper. First, using median filter for image enhancement and then using the threshold to binarize image according to its characteristics and removing interference target in the use of regional connectivity. Re-using map relationship to extract the image areas of coal and gangue. Extracting the gray feature through the analysis the characteristics of X-ray images of coal and gangue. Second, combining with the actual situation of belt conveyor gangue and CCD camera imaging principle on the basis of the traditional method of measuring height called calibration method and Proposing a method of measuring the height of gangue called proportionate calibration method. Third, using background elimination method to extract the areas of coal and gangue and adopting the proposed improved edge detection operator in the paper to detect the edge of the area. After that, using edge tracking algorithm to track edge detection, get the edge coordinatesand and obtain the height values of coal and gangue. Finally, Analyzing the basic principle of image recognition and drawing curves on two eigenvalues which are gradation and height by e xponential curve fitting to get representatives of the coal and gangue different curves. Then making the different curves as a basis for recognition.The experimental results show that the automatic recognition algorithm of coal and gangue based on X-ray image processing in the paper can accurately identify coal and gangue. Not only the recognition rate is almost 100%, but also having faster processing and recognizing. It has high stability and reliability.