Search results

This paper addresses the most relevant aspects of lossy image coding techniques, and presents an evaluation study on this subject, using several transforms and different methods in the transform domain. We developed different transform based image coders/decoders (codecs) using different transforms, such as the discrete cosine transform, the discrete wavelet transform and the S transform. Besides JPEG Baseline, we also use other techniques and methods in the transform domain such as a DWT based JPEG-like (JPEG DWT), a JPEG DWT with visual threshold (JPEG-VT), a JPEG–like coder based on the ST, and an EZW coder. The codecs were programmed in MATLAB™, using custom and built-in functions. The structures of the codecs are presented, also as some experimental results which allow us evaluate them, and support this study.

This paper presents a comparative study on image coders using the discrete cosine transform and the discrete wavelet transform, addressing in particular the JPEG-like and the EZW based image coders. We compare and evaluate the JPEG Baseline Image coder (using the DCT), a JPEG-like image coder based on the DWT, and an EZW coder, which is also based on the DWT. Taking into account the differences on the architecture of the coders, we compare them by using the same test images, for approximately the same conditions and characteristics, namely the approximately equal compression ratio.