Framelet Transform with DCT for Image Compression

Authors

  • Nora H. Sultan

Abstract

In this paper, an image compression using DCT technique and Framelet transform has introduced. This
technique is based on color and gray image compression. It uses Hybrid between Framelet Transform
and Discrete Cosine Transform to get higher quality than Joint Photographic Experts Group (JPEG),
JPEG2000 and Discrete Wavelet Transform (DWT). Good compression ratio has realized at first using
three levels of Framelet Transform (FLT) to obtain LL3 sub-band. Then, quantization and eliminating
zeros from each sub-band were employed. Then, compressing each sub-band by Arithmetic Coding
was done. Finally, compressing LL3 sub-band via DCT and then quantizing this sub-band and
compressing it by Arithmetic Coding have were used. The decompression algorithm involves Inverse
DCT to get decoded LL3. The inverse FLT was applied for getting a decompressed image. The adopted
algorithm has been tested on gray and color images and compared with JPEG and JPEG2000. The
projected algorithm gets the finest possible solution for a tradeoff amid compression ratio (size of
image) and quality of compressed image (PSNR). The simulation has been done by means of
MATLAB simulator.

Published

2020-12-04

Issue

Section

Articles