Tree-structured product-codebook vector quantization

This paper presents a new compression technique based on vector quantization. The proposed encoding algorithm uses a product codebook in which both component codebooks are tree-structured. This allows us to obtain low-complexity progressive encoding of large vectors. We propose a simple, greedy, procedure for the codebook design and assess its performance in the case of gray-scale images for gain-shape product VQ. Simulation results show that the proposed technique exhibits a significant performance gain at very low bit rates, while assuring a satisfactory image quality