Preservation of interlaced patterns in encoding video signals using side match vector quantization

Vector quantization (VQ) is a popular class of image encoding techniques which finds valuable applications in developing economical, low bit-rate video conferencing and multimedia products for the consumer's market. A new algorithm known as the side match vector quantization (SMVQ) had been proposed to enhance the visual quality of traditional VQ techniques by reducing the blocking effect based on local edge information. We show that despite the success of SMVQ in encoding still pictures, it does not perform equally well with interlaced video signals which contain motion contents. To overcome this problem, an algorithm known as the hybrid side match vector quantization (HSMVQ) algorithm which enable existing SMVQ to be applied to encode interlaced video signals with improved visual quality is developed and presented. Typical results obtained with HSMVQ, as compared to SMVQ are presented to demonstrate the feasibility of the approach