An Efficient Boundary Detection and Image Segmentation Method Based on Perceptual Organization
SRKIT, Ch Sambasivarao, V Naganjaneyulu · International Journal of Computer Trends and Technology · 2014
In this paper, we presents a novel method for detecting the boundaries of the object in outdoor images by using most common properties of the images such as perceptual organization laws. Here the proposed segmentation scheme is based on perceptual organization and background recognition. This paper mainly concentrates to recognize the structurally challenging objects, which is generally combination of several constituent parts. Our new proposed method based on perceptual organization model can efficiently recognize the non-accidental relationships, which are perfectly structured from the constituent parts of the strictly structured objects. The simulation results of this paper show that the efficient and accurate image segmentation by using perceptual organization models.