Volume : 4, Issue : 7, July - 2015

OBJECT LOCALIZATION USING FEATURE MATCHING IN CLUTTERED SCENES

Sneha Patil, Prof. N. C. Patil

Abstract :

<p>In many applications of computer vision, pattern recognition And medical image analysis, one common procedure Is to match two or more point sets, and nonrigid point set Matching is particularly difficult because the possible nonrigid Deformation of the model shape is numerous. In practice, the scene is often contaminated by clutters, making the point Set matching problem more complicated. In this paper, we focus on how to locate a deformable shape in cluttered scenes under the no rigid point set matching framework. The shape May undergo arbitrary translational and rotational changes, and It may be no rigidly deformed and corrupted by clutters.To address the problem of rotation invariant no rigid point set matching, they proposed two methods for shape representation. The shape context (SC) feature descriptor was used and we constructed graphs on point sets where edges are used to determine the orientations of SCs. This enables the proposed methods rotation invariant.The goal of this project is to automatically detect and recognize some objects in an image by using putative point matching algorithm.</p>