A Color Facial Image Segmentation using Bit Plane Slicing and Block Truncation Coding Techniques

Abstract

Facial image segmentation is vital and complementary in facial compression systems. IT generally represents the first step before using any coding techniques, but mostly utilizes traditional segmentation techniques of grayscale image base due to their simplicity and popularity. This paper introduces a new facial segmented image of the color base that exploits the spectral and spatial image information using differencing coding, and compression techniques of Bit Plane Slicing (BPS) and Block Truncation Coding (BTC), respectively. The tested results show that this technique is fast as it achieved about 0.3619 second on average and efficiently segmented the facial image automatically to nearly quarter the original image size.