Hybrid vector quantization

Natural images can be segmented into regions with widely varying perceptual importance. There are three types of regions in a typical image. The region in which the contrast between objects and background is high is categorized as the edge. Due to the high contrast, human eyes will naturally pay much attention to the edges. Thus, edge regions are very important for human perception. The second type of region is the smooth area, in which the brightness of neighboring pixels changes slowly. The third type of region is the texture part. Pixels in the texture region change slower than the edge region but faster than the smooth region. Typically, in a photograph such as ' 'Lena' ' [see Fig. 5(a)], the boundary of the face is categorized as the edge part, the cheek is the smooth part, and the hair is viewed as the texture part. We propose a new and simple method for classifying the edge and smooth parts; and then we encode the image using block coding with discrete cosine transform/vector quantization (DCT/VQ)' and the modified block truncation coding/vector quantization (BTCIVQ) derived from BTC/VQ.2 In our coding algorithm, we use block coding and, according to the eye model, treat the texture part as a composition of some smooth blocks. We then devote our attention Abstract. Vector quantization (VQ) is a powerful technique for low-bit-rate image coding. However, initial studies of image coding with VQ have revealed that VQ causes degradations, most notably around edges. Moreover, the computational complexity is high. Although a few algorithms have been developed to reduce edge degradation, such as block truncation coding (BTC) with VQ (BTCNQ) or classified vector quantization, their compression ratios are not satisfactory. Discrete cosine transformation with VQ (DCTNQ) has been applied to image compression, showing a high compression ratio, but the edge degradation problem still exists. We present an image compression algorithm that takes advantage of the merits of DCTNQ and BTCNQ to achieve a high-quality and low-bit-rate compression of images. High quality images can be achieved at rates of 0.34 to 0.46 bit/pixel.