【摘要】 EBCOT作为一种新的嵌入式图像压缩算法,比EZW和SPIHT算法具有更多的优越性,但它在具体实现中还尚有不清楚之处。作者针对"显著性标记"在位平面间及其内部邻域信息的传递,19种上下文索引值的确定原则及对MQ编码器输入信号的影响,以及MQ编码中状态转移等问题的不足,明确阐述了"显著性标记"对嵌入式和编码通道的重要性表示方式,19种上下文索引值与编码原语的对应关系及其在MQ编码器的输入信息中的获取途径,以及MQ编码中由概率分布模型和状态转移表引起的状态转换之实现方法。最后对其性能进行了分析。更多还原

【Abstract】 EBCOT as a new embedded image compression algorithm, have more advantages than EZW and SPIHT algorithms, but there are still problems. In this paper, considering the shortage as "significance label" change between bit-planes and its internal neighborhood, established principles of 19 kinds of context index and the influence to the MQ-encoder input signal, and status transfer in MQ coding, we present the express of "significance label" to embedded and encode pass, correspdence between 19 kinds of context index and coding language,and the MQ-encoder input signal access, as well as method causing by the probability distribution model and the state transition table. We analyse its performance at last.更多还原