Analysis of Shape Based Image Retrieval Using Different Wavelets Transforms

The shape of an obj ect is one of a defining factor in image processing. Here we will be focusing on image retrieval techniques relevant to shape features. The similarity within these shape features are calculated using wavelets transforms, principal components using Singular Value Decomposition(SVD) and clustering via K-Means.

I. Introduction

Visuals play a vital role in daily mundane life and they are conveying its message most of the time. Visuals are captured by us in form of images. There is deluge of images on web and it is natural for us to find similarity between images. This natural phenomenon gave rise to Internet Computer Vision or Content Based Image Retrieval system (CBIR). Any image contents are basically its features like color, shape and texture of an obj ect in that image or altogether as described by Rui et al., in [1]. In this paper we are stressing on shape based features.