Introduction - If you have any usage issues, please Google them yourself
Image compression is one of the prominent topics in image processing that plays a very important role in
reducing image size for real-time transmission and storage. Many of the standards recommend the use of DWT
for image compression. The computational complexity of DWT imposes a major challenge for the real-time use
of DWT-based image compression algorithms. In this paper, we propose a modified lifting scheme for
computing the approximation and detailed coefficients of DWT. The modified equations use, right shift
operators and 6-bit multipliers. The hierarchy levels in computation are reduced to one thereby minimizing the
delay and increasing throughput. The design implemented on Virtex-5 FPGA operates at 180 MHz and
consumes less than 1W of power. The design occupies less than 1 of the LUT resources on FPGA. The
architecture developed is suitable for real-time image processing on FPGA platform.