Pre-processing of Froth Image of Coal Flotation Based on Weighted Fuzzy C-Mean Clustering by One-Dimensional Histogram

Weighted Fuzzy C-Mean clustering algorithm by one-dimensional histogram can classify accurately pixels of coal flotation froth images into three kinds such as bubble vertex, the bubble surface and the background to increase the contrast of bubble edge and background, accurately positioning the bubble region for its fast convergence speed, real-time characteristics. After de-noised based on morphological filtering method of reconstruction opening and closing, clustered image can not only remove noises but retain more image detail and reflect accurately the true image information. So over-segmentation and under-segmentation will be avoided so as to be segmented more accurate when image clustered by these algorithms is processed using watershed segmentation based tags.