On the selection of appropriate wavelet filters for visual sensor networks

With the emergence of visual sensor networks (VSN), low power wavelet-based coder (WBC) is becoming increasingly mandatory. This makes the selection of the appropriate wavelet, among many competitors, not an easy task. In our context, the appropriate wavelet is that one which dissipates low energy during image decomposition, while having an adequate quality of the reconstructed image at the reception. In this paper, a comparative study is investigated between different wavelet filters. Two versions of DWT implementation are considered following their emergence: the classical convolutional-based wavelets and the relatively new lifting-based wavelets.