Seperating Touching Particles: A Concavity-Based Method Using the Area Ratio of a Circular Mask

• Published in: IEEE systems, man, and cybernetics magazine man, and cybernetics magazine, 2018-04, Vol.4 (2), p.24-32
• Main Authors: Wang, Wencheng, Yuan, Xiaohui, Wu, Xiaojin, Ji, Tao, Feng, Lin
• Format: Magazinearticle
• Language: English
• Subjects: Corner detection; Cybernetics; Image edge detection; Image segmentation; Particle measurements; Shape analysis; Transforms
• ISSN: 2380-1298; 2333-942X
• DOI: 10.1109/MSMC.2018.2794559

To address the difficulty of separating touching particles in an image, a method based on the area ratio of a circular mask is proposed in this article, and a framework consisting of image segmentation, concave point detection, point-pair matching, and the shape estimation of particles is implemented. A given color image is first preprocessed to obtain a binary image, then a corner detection algorithm is adopted to obtain a rough estimation of potential concave points, following which, the area ratio method is used to refine segmentation points qualified. The segmentation points obtained are matched to each pair by gray projection according to a certain evaluation criteria. Finally, each pair of points and one of their adjacent pixels on the border are used to estimate the status of each particle before adhesion, and the contour of each particle is generated using the curve fitting method. Experimental results show that the proposed method is consistent with human visual perception and can be used in the practice.