This paper presents a lossless index-coding algorithm for image vector quantization. Using the concept of dynamic tree coding, the method can encode the output index map of a memoryless vector quantizer efficiently without introducing any extra coding distortion. Experiment results show that it achieves better compression efficiency than other lossless index coding schemes. The computation complexity of the method is quite low and its memory requirement is small. Hence, it can be easily implemented by hardware and applied to many real-time applications.
All Science Journal Classification (ASJC) codes
- Media Technology
- Electrical and Electronic Engineering