Splitting ore from X-ray image based on improved robust concave-point algorithm.

Abstract

Image segmentation is a key part of ore separation process based on X-ray images, and its segmentation result directly affects the accuracy of ore classification. In the field of ore production, the conventional segmentation method is difficult to meet the requirements of real-time, robustness and accuracy during ore segmentation process. In order to solve the above problems, this article proposes an ore segmentation method dealing with pseudo-dual-energy X-ray image which is composed of contour extraction module, concave point detection module and concave point matching module. In the contour extraction module, the image is firstly cut into two parts with high and low energy, then the adaptive threshold is used to obtain the ore binary image. After filtering and morphological operation, the image contour is obtained from the binary image. Concave point detection module uses vector to detect concave points on contour. As the main contribution of this article, the concave point matching module can remove the influence of boundary interference concave points by drawing the auxiliary line and judging the relative position of auxiliary line and ore contour. With the matching concave points connected, the whole ore segmentation is completed. In order to verify the effectiveness of this method, a comparative experiment was conducted between the proposed method and conventional segmentation method using X-ray images of antimony ore as data samples. The result of industrial experiment shows that the proposed intelligent segmentation method can remove the interference of pseudo concave points on the contour, achieve accuracy segmentation result, and satisfy the requirements of processing X-ray image of ore.