Showing papers by "Jian Zhou published in 2009"
14 Aug 2009
TL;DR: This paper investigates the novel effective segmentation method SCCPSO by discussing several key points such as membership function, initialization, pre-selection, and boundary conditions, and applies it in a CV-based inspection system.
Abstract: As a critical unit of computer vision (CV) based applications, image segmentation is quite worth studying. Hybrid method of spatial credibilistic clustering and particle swarm optimization (SCCPSO) is a novel effective segmentation method. It's proved to produce better results than other common methods. In this paper, SCCPSO is further investigated by discussing several key points such as membership function, initialization, pre-selection, and boundary conditions. Then the modified SCCPSO is put forth and applied in a CV-based inspection system to show its effectivity and better performance. The proposed method can be also used in other CV-based applications.
4 citations