在资源受限的无线多媒体传感器网络(WMSNs)中进行图像编码和传输需要综合考虑能量消耗、压缩率和图像质量三者之间平衡的图像编码方案。对基于离散小波变换的图像编码算法的能耗进行建模分析,提出了一种适用于WMSNs的能量有效的JPEG2000图像编码算法,根据网络条件和图像质量的限制,使用查找表来选择适当的量子化层级和小波变换层级以减少能量消耗。并采用半可靠的方案进行图像传输,节点根据剩余能量和数据优先级来决定转发或丢弃。仿真实验结果表明:所提出的方法能够在保证所要求图像质量的情况下,有效地降低无线传感器节点的计算和通信能耗。
In resource-constrained wireless multimedia sensor networks (WMSNs) ,image coding and transmission must respect the trade-off among energy consumption, compression ratio and image quality. Energy consumption model of discrete wavelet transform(DWT) based image compression algorithm is developed. An energy efficient JPEG 2000 image compression algorithm is adapted to WMSNs. The algorithm determines the parameter of quantization and the wavelet image compression transform level to meet the image quality requirement and channel condition. Semi-reliable transmission enables packet forward or discarding by intermediate nodes according to residual energy and data priority. Simulation results show that the proposed image compression scheme provides a graceful trade-off between the reconstructed images quality and the total energy consumption of computation and communication of sensor nodes.