A Low Complexity 2D Hidden Markov Model in Application to Image Segmentation
Yu Lu①②; Wu Le-nan①; Xie Jun③
①Department of Radio Engineering, Southeast University, Nanjing 210096, China;②Institute of Communications Engineering, PLA University of Science and Technology, Nanjing 210007, China;③Institute of Command Automation, PLA University of Science and Technology, Nanjing 210007, China
Abstract:The assumption of conditional independence in the relationship between adjacent blocks has been proposed by others to reduce the complexity of 2D HMM. In this paper, a more general 2D HMM relaxing this assumption is proposed. More general recursive forms of the forward and the backward algorithms are derived. And the model provides more flexibility by adjusting the weight between horizontal and vertical information. The application to image segmentation verifies the effectiveness of the model.