The Research Of Image Process And Detect Coal Level Underground Coal Bunker

With the modernization of coal development, in order to product and manage automaticly, we urgently need continuously monitor and control coal bunker coal level at any time to ensure various production processes to be safe, efficient, and to improve productivity and automation management level.Coal level measurement in the world coal industry has long been the problem has not been completely resolved. Over the years, people have tried a variety of material level detection methods. The condition of underground bunkers are bad, the results have not been very good. In this paper, use image processing and pattern recognition technology to research coal level detection. According to the features of underground bunkers, firstly using two laser light source imaging on the coal surface, according to distance between two laser points to detect the coal level, this method is avoid the problem that unsmooth coal face lead to be irregular image and difficult to detect coal level, researching suitable image acquisition system, processing, segmentation and recognition. Image acquisition system mainly introduces hardware and figure about the basic principle, its hardware and equipment include laser lights, CCD camera, image acquisition card, computer and shell explosion. In the image processing, mainly introduces the feature of general image noise and noise of image from underground bunkers and basic image filter and image enhanced techniques, including Smooth filter, Median filter, Maximum filter, Wiener filter, Gradient operator, Laplacian operator. Analysis of a variety of filter and enhance technologies about the features and applications. Image segmentation, mainly for edge detection algorithm, Roberts operator, Laplacian - Gaussian operator. Canny operator, Sobel operator, Prewitt operator, Gray thresholding segmentation. algorithm, regional growth algorithm. Image Recognition technology, this paper will focus on statistical pattern recognition, fuzzy pattern recognition and pattern recognition methods for structural studies. According coal bunker imaging features, firstly use the Maximum filter for coal bunker digital image processing, Sobel operator is used to the image segmentation. Finally, use the minimum distance methods to detect coal level, the experimental results show the proposed method is effective, quickly and accurately to detect coal level, the experiment achieved good results.