Content Based Image Retrieval Method : Based On Low Level Features

ISBN-13: 9783659266065 / Angielski / Miękka / 2012 / 96 str.

Smita Das; Dwijen Rudrapal

Opis

In this book, a study and analysis is done to improve the Content Based Image Retrieval method which uses the traditional low level features of images. Supported experimental results for a variety of classes of images depicted higher retrieval accuracy and lower computation speed. This method has faster retrieval speed than other methods due to its multiple features and low feature vector dimensions. The following are some key features of this book - Analysis of visual contents of an image is prepared by vast literature survey which has exposed that Color and Texture based features are most effective in content based image retrieval systems. Proper use of Color Spaces has improved the accuracy of the system. Special attention has been given to balance the article with mathematical models, notations and descriptions. As compared to single low level feature based image retrieval, multiple features like color and texture together gives better accuracy and high retrieval rate. Every diagram in this book is demonstrated accurately for clarification of the idea.

O autorze

Smita Das, having more than 6.5 years working experience in the field of Teaching and R & D with a good number of Research articles, is presently working as Assistant Professor in CSE Dept. in National Institute of Technology Agartala, Tripura, INDIA. Her current field of research interest is Content Based Image Retrieval.