摘要
SPIHT(set partitioning in hierarchical tree)是一种简单、高效的小波零树图像编码算法。该文针对 SPIHT算法存储空间需求大 ,不利于硬件实现的缺点 ,引入“误差位数”、“绝对零点”和“绝对零树”的概念 ,提出了一种新的低存储量的图像编码算法 ,并利用 DSP评估板 (EVM)对新算法进行了验证。大量实践证明 ,新算法有效地降低了算法实现所需的存储空间 ,减少了时间消耗 ,易于硬件的实现。而且 ,新算法重建图像的峰值信噪比 (PSNR)指标对比
SPIHT (set partitioning in hierarchical tree) algorithm is a wavelet and zerotree image coding algorithm known for its simplicity and efficiency. However, SPIHT's high memory requirement is an obstacle to hardware implementation. This paper provides a new low memory image zerotree coding algorithm using three concepts 'error digit', 'absolute zero coefficient' and 'absolute zerotree'. The performance of the new algorithm was tested using the DSP Evaluation Module (EVM). Through extensive experiments, the results show that it drastically reduces both the memory requirement and the time consumption, with only a minor reduction in PSNR values when compared with those obtained by the SPIHT codec.
出处
《清华大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2001年第9期59-62,66,共5页
Journal of Tsinghua University(Science and Technology)
关键词
图像编码
小波变换
零树量化
SPIHT
PSNR
image coding
wavelet transform
zerotree quantization
SPIHT (set partitioning in hierarchical tree)
PSNR (peak signal to noise ratio)
