Abstract

In this paper, a fast image processing was proposed. It ensures energy efficiency and the extension of both the lifetime and the proper functioning of the network. It is a filtered zonal discrete cosine transform that allows and optimizes an effective adjustment of the trade-off between power consumption and image distortion. This is a remarkable energy saving method, in this kind of networks. It is applied throughout the chain of transmission and decompression of the image. It makes it possible to integrate a fast and a filtered zonal discrete cosine transform. This proposal dramatically improves the indicated method. The insertion of the frequency filters in this chain has greatly reduced the coefficients to be calculated and to be coded in each block. This new method ensures the fast transfer of images, decreases more the energy consumption of sensors and maintains a long network lifetime. This proposal seems to us very satisfactory as shown by the experimental results provided here.

Key words: Energy saving, fast zonal discrete cosine transform, filtered fast zonal discrete cosine transform, image compression, wireless vision sensors network, zonal coding.

Abbreviation

DCT, Discrete cosine transform; WSN, wireless sensors network; WCN, wireless cameras network; MSE, mean square error; DC(.,.), discrete coefficient; FZ_DCT, fast zonal discrete cosine transform; FFZ_DCT, filtered fast zonal discrete cosine transform; RLE, run length encoding; LLM, Loeffler-Ligtenberg-Moschytz; LEACH, low-energy adaptive clustering hierarchy; SSIM, structural similarity index measure; MSSIM, mean structural similarity index measure.