A novel region-based image segmentation method using SVM and D-S evidence theory
Shuai Li, Tao Leí, Xiaojun Jing, Songlin Sun, Yueming Lu, Chenglin Zhao, Na Chen · 2013
Region-based image segmentation is an important preprocessing step for high-level computer vision tasks. This paper presents a novel approach to image partition into regions that reflect the objects in a scene. It explores the feasibility of utilizing Gray Level Co-occurrence Matrix (GLCM) and RIQ color feature of regions to improve the segmentation results produced by Recursive Shortest Spanning Tree (RSST) algorithm. Combination of Support Vector Machine (SVM) and Dempster-Shafer (D-S) theory is applied to the field of region merging. In the proposed algorithm, SVM is utilized as the identifier, and Basic Belief Assignment (BBA) function is constructed accordingly. Fused BBAs are obtained by applying the D-S evidence theory to the outputs of the identifiers. The experimental results show that the proposed method provides higher accuracy and stability when compared with the original RSST segmentation algorithm.