Color, Texture and Shape Based Relevance Feedback Image Retrieval for Web Applications

C Mythili¹

¹Assistant Professor, University College of Engineering, Nagercoil

Abstract

Content Based Image Retrieval (CBIR) plays a very vital role in the retrieval of relevant images queried by a user for web applications. This paper describes an efficient Relevance Feedback (RF) mechanism for CBIR. The image retrieval is performed based on the retrieval of color, texture and shape features. The RGB input image is converted to three different color spaces like HSV, YCbCr and CIE Lab color spaces. The statistical color features such as mean, variance and skewness are extracted from the above color space. The local texture information from each pixel determines the fuzzy texture unit. The shape feature is obtained through the segmentation process. Then these feature dimensions such as 'color- 27', 'texture- 2020', and 'shape- 128' are combined to form feature vector. This feature vector of the query image is compared with the feature vector of the images in the Benchmark databases such as General database, Medical database, Berkeley databases, COREL database and COIL database using Euclidean distance. Finally RF is incorporated to improve the accuracy of the image retrieval. Results have proved that the Precision and Recall are much improved through the increase of iteration of RF.

Keywords

• Color feature, Texture feature, Shape feature, Relevance Feedback

Author's Profile

Author profile can be generated and linked through our partners World Book of Researchers. To include your profile online Click Here. After it is approved, please email to edlib @ asdf.res.in to create a link with all the papers.

• C Mythili : Profile

e-AID

Cite this Article as Follows

C Mythili. Color, Texture and Shape Based Relevance Feedback Image Retrieval for Web Applications. International Conference on Inter Disciplinary Research in Engineering and Technology 2014. Vol. 1. Chennai: Association of Scientists, Developers and Faculties, 2014. 1-6. Print.