A Low-Complexity Interleaved Image Wavelet Transform Architecture for a Visual Sensor Node

We present a low-complexity architecture to perform a wavelet subband decomposition on an image captured from a visual sensor. The input into the architecture is a dynamic stream of image data from the visual sensor. The subband coefficients are computed on-the-fly as the data flow through the architecture. The output from the architecture is a stream of interleaved subband coefficients. A 160times120 image pixels architecture to perform three scales of subband decomposition has been synthesized and implemented on a Xilinx FPGA utilizing a chip area of 940 Slices