Block Truncation Coding for color Images Using Vector Quantization

Abstract: The method of block truncation coding (BTC) was originally proposed by Delp and Mitchell and later extended to color images by others. The idea is to attempt to retain to important visual features while discarding details which are not to be noticeable to human observers. In this paper, further explore the block truncation coding method was done by minimizing the within group variance measure proposed by Otsu and the information distance suggested by kullback to divide every 4x4 subimages into two classes, and an intuitive vector quantizer to further compress the coded output. As a result of the combined application of BTC and vector quantization methods, we get better bit rates (i.e., 1 bits per pixel) for the test images used in experiments without any significant perceivable errors in their appearance. First, we divide a color image into 4x4 small nonoverlapping blocks. Otsu and Kullback distance measure is then used as an optiomal method to minimize the mean square error of classifying each pixel in a block into two classes and encode it in one-bit adaptive vector quantizer. After classification, for each 4x4 block, there is a bitmap corresponding to one-bit adaptive vecotor quantizer and a six-diminsional mean vector corresponding to each of the two classes. In the second part, the vector quantizer proposed by linde, Buzo and Gary (known as LBG) is used to compress the bit-map and mean vectors separately. This is a six-dimensional signal compression for the mean vectors and a binary compression for the bitmap. Vector quantization for these BTC outputs results in a reduction of the bit rate of the coder. By using BTC and vector quantization methods, we have obtained 1.0 bit/pixel compression result for color images of size 512x480 given with 8 bits/pixel and R, G, B specifications. The mean square error was also measureed as low as 0.07 without much deformation in the reconstructed images.