The USC Andrew and Erna Viterbi School of Engineering USC Signal and Image Processing Institute USC Ming Hsieh Department of Electrical Engineering University of Southern California

Technical Report USC-SIPI-251

“JPEG-based Image Compression Using Wavelet Transform”

by Kwo-Jyr Wong, Po-Yuen Cheng and C.-C. Jay Kuo

March 1994

The block DCT (Discrete Cosine Transform) plays an important role in the JPEG still image compression standard. In this research, we investigate the replacement of the block DCT in the JPEG with the fully-decomposed wavelet-packet transform (FWT) while keeping all other building blocks the same with the objective to improve the performance of the JPEG. Following the FWT, two postprocessing operations, i.e. frequency band swapping and frequency/spatial coefficient rearrangement, are implemented. The JPEG quantization and coding techniques are then applied. The resulting method gives better performance than the DCT-based JPEG scheme in the MSE (mean square error) measure as well as visual appearances.

To download the report in PDF format click here: USC-SIPI-251.pdf (1.0Mb)