DocumentCode :
3106755
Title :
Image compression and gamma correction using DCT
Author :
Mohta, Jyoti ; kant Pathak, Krishna
Author_Institution :
Electron. & Commun., RGPV, Bhopal, India
fYear :
2015
fDate :
25-27 Feb. 2015
Firstpage :
322
Lastpage :
326
Abstract :
In this paper an automatic image compression method is present. Image compression is done by modifying the gamma value of each pixels rather than applying uniform value of gamma like most of the exiting gamma correction methods follows. This method includes compression of digital images with the help of discrete cosine transform (DCT). To improve the performance of image compression, several encoding techniques are also used together with DCT. These techniques are Run length encoding and Differential pulse code modulation (RLE-DPCM). Both techniques are used separately on AC and DC coefficient of DCT of an image. In this algorithm, first the original image is divided into blocks where block size is also specified. Then these blocks are subjected to DCT. Through standard JPEG table each transformed block is also divided and then we set the coefficient of quantization table by scaling factor. The image performance is measured in terms of CR, MSE and PSNR.
Keywords :
data compression; differential pulse code modulation; discrete cosine transforms; gamma distribution; image coding; mean square error methods; AC coefficient; CR; DC coefficient; DCT; MSE; PSNR; RLE-DPCM; automatic image compression method; differential pulse code modulation; discrete cosine transform; encoding techniques; gamma correction methods; image block size; image pixels; performance improvement; performance measured; quantization table coefficient; run length encoding; scaling factor; standard JPEG table; uniform gamma value; Discrete cosine transforms; Finite element analysis; Image coding; Knowledge management; Market research; Quantization (signal); Transform coding; DCT; Image Compression; RLE-DPCM Encoding;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Futuristic Trends on Computational Analysis and Knowledge Management (ABLAZE), 2015 International Conference on
Conference_Location :
Noida
Print_ISBN :
978-1-4799-8432-9
Type :
conf
DOI :
10.1109/ABLAZE.2015.7155005
Filename :
7155005
Link To Document :
بازگشت