THE OPTIMAL QUANTIZATION MATRICES FOR JPEG IMAGE COMPRESSION FROM PSYCHOVISUAL THRESHOLD

The JPEG image compression method has been widely implemented in digital camera devices. The quantization process plays a primary role in JPEG image compression. The quantization process is used to determine the visibility threshold of the human visual system. The quantization tables are generated from a series psychovisual experiments from several angle points of experimental views. This paper proposes psychovisual threshold through quantitative experiments for JPEG image compression. This experiment investigates the psychovisual threshold based on the contribution of DCT coefficients on each frequency order to the reconstruction error. The average reconstruction error from incrementing DCT coefficient is investigated to produce a primitive psychovisual threshold. The psychovisual threshold is designed to give an optimal balance between quality of image reconstruction and compression rates. A psychovisual threshold is obtained to generate new quantization tables for JPEG image compression. The performance of new quantization tables from the psychovisual threshold are analyzed and compared to the existing default JPEG quantization tables. The experimental results show that the new quantization tables from the psychovisual threshold produce higher quality of image reconstruction at lower average bit-length of Huffman code than default JPEG quantization tables.

[1]  Joan L. Mitchell,et al.  JPEG: Still Image Data Compression Standard , 1992 .

[2]  D. Mclaren,et al.  Removal of subjective redundancy from DCT-coded images , 1991 .

[3]  Xiaolin Wu,et al.  A psychovisually tuned image codec , 2011, 2011 IEEE 13th International Workshop on Multimedia Signal Processing.

[4]  Ramakrishnan Mukundan,et al.  A Comparison of Discrete Orthogonal Basis Functions for Image Compression , 2004 .

[5]  Hugo Fastl,et al.  Psychoacoustics: Facts and Models , 1990 .

[6]  King Ngi Ngan,et al.  Adaptive cosine transform coding of images in perceptual domain , 1989, IEEE Trans. Acoust. Speech Signal Process..

[7]  David A. Huffman,et al.  A method for the construction of minimum-redundancy codes , 1952, Proceedings of the IRE.

[8]  Eero P. Simoncelli,et al.  Image quality assessment: from error visibility to structural similarity , 2004, IEEE Transactions on Image Processing.

[9]  En-Hui Yang,et al.  Joint Optimization of Run-Length Coding, Huffman Coding, and Quantization Table With Complete Baseline JPEG Decoder Compatibility , 2009, IEEE Transactions on Image Processing.

[10]  Weisi Lin,et al.  A Psychovisual Quality Metric in Free-Energy Principle , 2012, IEEE Transactions on Image Processing.

[11]  M. Mrak,et al.  Picture quality measures in image compression systems , 2003, The IEEE Region 8 EUROCON 2003. Computer as a Tool..

[12]  N. Ahmed,et al.  Discrete Cosine Transform , 1996 .

[13]  Alan C. Bovik,et al.  The Essential Guide to Image Processing , 2009, J. Electronic Imaging.

[14]  Somkait Udomhunsakul,et al.  Objective Performance of Compressed Image Quality Assessments , 2007 .

[15]  M. Guy,et al.  “ An Improved Detection Model for DCT Coefficient Quantization , 2007 .