摘要
介绍了一种基于贝叶斯估计的Context量化器设计方法.通过将自适应码长增量与贝叶斯估计中的先验概率估计进行结合,使得该量化器在不需要先验知识的情况下,既考虑到Context量化本身的特点,又使得编码后信源的自适应码长最短,最终保证了Context量化的自适应性.实验结果表明,该Context量化器设计方法能获得最优量化结果,达到设计目标.
In this paper, a new context quantization algorithm based on the Bayesian estimation is proposed and combined the increment of auto-adaptive code length with the prior conditional probability in Bayesian estimation to make this Context quantization without any prior conditions think about its own features and minimize the code length of the source and at last guarentee the Context quantiza- tion auto-adaptation. The result showed that this algorithm of Context quantization can reach the best result and the aim of designment.
出处
《昆明学院学报》
2013年第3期79-82,共4页
Journal of Kunming University
基金
昆明学院校级科学研究基金资助项目(XJL12003)
关键词
贝叶斯估计
Context量化
自适应码长
贝叶斯分类
Bayesian estimation
Context quantization
auto-adaptive code length
Bayesian classification
