Title :
Quantization scheme for JPEG baseline sequential encoding of still images
Author :
Paik, Hyung W. ; Khubchandani, Anil
Author_Institution :
Dept. of Electr. Eng., South Dakota Sch. of Mines & Technol., Rapid City, SD, USA
Abstract :
A quantization scheme for discrete cosine transform (DCT) coefficients in the Joint Photographic Experts Group´s (JPEGs) baseline sequential method for image compression is proposed. The DCT coefficients should be quantized to achieve maximum compression without degrading the visual image quality. The scheme is adaptive to the image characteristics and is statistical in nature. The results are evaluated in terms of compression, root-mean-square error, and subjective visual quality, 8-b/pixel monochrome images of size 512 × 512 have been compressed in the range 0.4-1 b/pixel with good to excellent quality
Keywords :
analogue-digital conversion; data compression; discrete cosine transforms; image coding; 262144 pixels; 512 pixels; JPEG baseline sequential encoding; Joint Photographic Experts Group´s; discrete cosine transform; image characteristics; image compression; monochrome images; quantization scheme; root-mean-square error; still images; subjective visual quality; visual image quality; Color; Decoding; Discrete cosine transforms; Entropy coding; Equations; Image coding; Image reconstruction; Pixel; Quantization; Transform coding;
Conference_Titel :
Circuits and Systems, 1992., Proceedings of the 35th Midwest Symposium on
Conference_Location :
Washington, DC
Print_ISBN :
0-7803-0510-8
DOI :
10.1109/MWSCAS.1992.271133
