A novel parallel architecture of lifting based 2D-discrete wavelet transform

This Image compression using Discrete Wavelet Transform has got application in many filed like biomedical, agriculture, digital video compression, wireless networks. In this paper we present a parallel-pipelined architecture for lifting based two dimensional DWT. We have used lifting based wavelet filter for image compression. Overall area has been reduced by eliminating line buffer and frame buffer from the design. By proper scheduling of the intermediate results this architecture avoid the need of transposition buffer. The architecture has a very small latency (a few cycles) and provide high throughput. By removing redundancy the hardware utilization efficiency has been maximized. The architecture is simulated on ISE simulator and hardware implemented on Artex board.