Image retrieval using Block Truncation Coding Extended to Color Clumps

Image retrieval is useful to retrieve images from outsized image databases, which can be beneficial to plenty of image supporting applications. Colors of an image are easier for extraction. Block Truncation Coding (BTC) prominently used with many variations. This paper proposed a novel Block Truncation Coding Extended to Color Clumps for image retrieval purpose. Total of 24 variations, using four clumps and six color spaces are experimented on image database having 1000 images. Experimental results have shown better performance in YCbCr color space followed by YUV and LUV. The best image retrieval is by Extended Block truncation coding using 8 color clumps in YCbCr color space.