DocumentCode :
3081906
Title :
Global DCT Domain Power Law Transformations in Image Enhancement Technique
Author :
Singh, Twinkle R. ; Roy, Sandip ; Manglem Singh, Kh
Author_Institution :
Dept. of IT, Assam Univ., Silchar, India
fYear :
2013
fDate :
24-26 Aug. 2013
Firstpage :
269
Lastpage :
273
Abstract :
Power Law Transformation (PLT) is a standard transformation function used in image enhancement techniques. PLT associates two constant parameters which can change an input value in different ways. The two PLT parameters can take different roles on different domains like spatial and transform domain. Discrete Cosine Transformation(DCT) is the popular transform domain used in image processing. Alpha Rooting (AR) and DCT block domain techniques are popular image contrast enhancement techniques which apply PLT on DCT domain. They are limited to edge contrast enhancement only. This paper describes an image enhancement technique which applies PLT on a global DCT domain like AR. PLT is applied only on AC coefficients while AR applies to all the DCT coefficients. This new technique can enhance global contrast, edge contrast and brightness of an image through a single function with three parameters. The results of this technique are compared with other techniques, applied on transformed domain for quality and quantity measure. This technique outperforms than the other techniques reported so far.
Keywords :
discrete cosine transforms; edge detection; image enhancement; AC coefficients; AR; DCT block domain techniques; DCT coefficients; PLT parameters; alpha rooting; discrete cosine transformation; edge contrast enhancement; global DCT domain power law transformations; global contrast; image brightness; image contrast enhancement techniques; image processing; quality measure; quantity measure; spatial domain; standard transformation function; transform domain; Brightness; Discrete cosine transforms; Frequency-domain analysis; Image edge detection; Image enhancement; Image reconstruction; Compress domain; Contrast; DCT; Enhancement; Power-law transformation; Sharpness; Smoothness;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computational and Business Intelligence (ISCBI), 2013 International Symposium on
Conference_Location :
New Delhi
Print_ISBN :
978-0-7695-5066-4
Type :
conf
DOI :
10.1109/ISCBI.2013.61
Filename :
6724366
Link To Document :
بازگشت