Design and Implementation of IP Core for Contourlet-Based Image Compression

Abstract

This paper presents a Contourlet-based image compression algorithm and its hardware IP core design. The proposed algorithm is based on the Set Partitioning in Hierarchical Trees (SPIHT) coding method in Contourlet domain. Three new ideas for this algorithm are proposed: finding significant subbands by analyzing and comparing the distribution of coefficients in both Contourlet and Wavelet subbands; rearranging the order of these subbands based on their significant level; coding more significant subbands first by more bits to reduce compression loss. Thanks to these improvements, the proposed algorithm outperforms the other methods using SPIHT in Contourlet or Wavelet domain. In order to increase the processing time of the proposed algorithm, its hardware IP core is designed and verified on Cyclone IV FPGA. The architecture of the IP core supports Avalon bus interface and is able to be easily integrated into a system-on-chip (SoC). The complete verification of the IP core is described to demonstrate its correctness and applications in image compression.