论文部分内容阅读
提出一种基于双重截断的JPEG 2000码率控制算法,并给出其VLSI架构。双重截断包括基于熵估计的码率预分配的粗略截断和易于硬件实现的Tier-2快速截断。首先根据各个码块的熵估计值在所有码块熵估计值总和中所占的比例,提前分配码率,实时控制Tier-1编码部分对码流的第一次截断;再利用一种Tier-2快速码流截断法寻找最优截断点,进行码流的第二次精细截断。实验结果表明,通过双重截断所得的压缩图像,与标准PCRD-opt压缩图像的峰值信噪比差值最大不超过0.5dB。
A JPEG 2000 rate control algorithm based on double truncation is proposed and its VLSI architecture is given. Dual truncation includes coarse truncation of rate-based pre-assignment based on entropy estimation and fast Tier-2 truncation of hardware-easy implementation. First, according to the proportion of each block’s entropy estimation value in the sum of the estimated values of all the block entropies, the code rate is allocated in advance to control the first truncation of the code stream in the Tier-1 coding part in real time. 2 fast stream truncation method Find the best truncation point, the second stream of fine truncation. The experimental results show that the maximum SNR difference with the standard PCRD-opt compressed image does not exceed 0.5dB at most by the double truncation of the compressed image.