基于改进SPIHT的传感网络低复杂度图像压缩算法

Low complexity image compression algorithm for sensor networks based on improved SPIHT

为有效降低传感网络图像压缩算法的计算复杂度,提高偏远区域无线传感网络的监测与传输性能,提出一种改进低复杂度多级树集合分裂(Set Partitioning in Hierarchical Trees,SPIHT)的传感网络图像压缩算法。该算法在图像小波分解过程中,利用提出的两种池化决策裁剪和优化高频系数,解决了SPIHT在重要性系数选择方面计算复杂度高的问题。通过实验对比分析,对于同类型和不同类型特征的图像,在保证重构图像清晰度的条件下,提出算法在编码、解码时间上分别比SPIHT算法平均减少38.47%和44.11%,有效降低了计算复杂度,提升了传感网络监测与传输能效。

In order to decrease the computational complexity of image compression algorithms and improve the monitoring and transmitting performance of wireless sensor networks in remote areas,an improved low-complexity image compression algorithm based on Set Partitioning in Hierarchical Trees(SPIHT)is proposed.This algorithm utilized two pooling strategies to cut and optimize the part of high-frequency coefficients in image wavelet decomposition process,which solved the problem of high computational complexity in the process of selecting importance coefficients in SPIHT.Through experimental and comparative analysis,the quality of reconstructed images of the same and different image types was ensured.Compared with SPIHT,the encoding and decoding time of the proposed algorithm was reduced by 38.47%and 44.11%respectively.The proposed algorithm can effectively decrease the complexity of SPIHT algorithm,and upgrade the efficiency of sensor network monitoring and transmission.