基于稀疏变换的图像压缩方法的研究

针对无线多媒体传感器网络中节点计算能力、存储空间和能量等资源受限问题,提出了联合稀疏变换的分布式压缩感知视频压缩方法。该方法对传感器采集到的图像序列先进行分布式视频编码,再利用联合稀疏变换对图像进行稀疏表示,其中联合稀疏变换方法结合了冗余字典和正交基稀疏变换的优点,比冗余字典稀疏表示得到的图像更稀疏,压缩效率更高。仿真结果表明,该方法有效减少了编码和传输的数据量,降低了图像压缩传输时间。

Optimization of Image Compression Method Based on Sparse Transform

The key challenges of the wireless multimedia sensor network are the low power,computational capabilities and the limited energy of the sensor nodes. We propose a new method- the Joint Sparse Transform. It combines with redundant dictionary and orthogonal matrix sparse transform.Compared to the redundant dictionary,the Joint Sparse Transform achieves higher compression efficiency. The simulation results show that,the optimization theory of sparse transform is effective to reduce the amount of data of encoding and transmission and improve the network performance.