文章基本信息

标题：Improving Performance of Multileveled BTC Based CBIR Using Sundry Color Spaces
本地全文：下载
作者：Dr. H.B.Kekre ; Mr. Sudeep D.Thepade ; Mr. Srikant Sanas 等
期刊名称：International Journal of Image Processing (IJIP)
电子版ISSN：1985-2304
出版年度：2011
卷号：4
期号：6
页码：620-630
出版社：Computer Science Journals
摘要：The paper presents an extension of content based image retrieval (CBIR) techniques based on multilevel Block Truncation Coding (BTC) using nine sundry color spaces. Block truncation coding based features is one of the CBIR methods proposed using color features of image. The approach basically considers red, green and blue planes of an image to compute feature vector. This BTC based CBIR can be extended as multileveled BTC for performance improvement in image retrieval. The paper extends the multileveled BTC using RGB color space to other nine color spaces. The CBIR techniques like BTC Level-1, BTC Level-2, BTC Level-3 and BTC Level-4 are applied using various color spaces to analyze and compare their performances. The CBIR techniques are tested on generic image database of 1000 images spread across 11 categories. For each CBIR technique, 55 queries (5 per category) are fired on extended Wang generic image database to compute average precision and recall for all queries. The results have shown the performance improvement (ie., higher precision and recall values) with BTC-CBIR methods using luminance-chrominance color spaces (YCgCb, Kekre's LUV, YUV, YIQ, YCbCr) as compared to non-luminance (RGB, HSI, HSV, rgb , XYZ) Color spaces. The performance of multileveled BTC-CBIR increases gradually with increase in level up to certain extent (Level 3) and then increases slightly due to voids being created at higher levels. In all levels of BTC Kekre's LUV color space gives best performance
关键词：Gender Recognition; Feature Extraction; Edges