Title :
Energy efficient, architectural reconfiguring DCT implementation of JPEG images using vector scaling
Author :
Jessintha, D. ; Reghu, C. ; Raj, A. Vimal
Author_Institution :
Dept. of Electron. & Commun. Eng., Anna Univ., Chennai, India
Abstract :
This paper presents the DCT implementation of JPEG images using discrete cosine transform (DCT). The proposed method aims at reducing the power consumption which by application of vector scaling on DCT. This will be very useful for image compression and transmission applications. Using the character of energy distribution of DCT matrix after 2D-DCT operation, the best DCT basis functions are selected which is capable of achieving considerable power reduction with minimum image quality degradation. The proposed method can efficiently trade-off image quality and power consumption. An algorithm with vector scaled values for DCT coefficients, is also proposed which reduces the computational energy based on the sensitivity differences of 64 DCT coefficients.
Keywords :
data compression; discrete cosine transforms; image coding; reconfigurable architectures; 2D-DCT operation; JPEG images; architectural reconfiguring DCT implementation; discrete cosine transform; image compression; power consumption; vector scaling; Complexity theory; Computer architecture; Degradation; Discrete cosine transforms; Image quality; Sensitivity; Transform coding; Discrete cosine transform (DCT); alphabet selection; power saving implementation; trade off algorithm; vector scaling;
Conference_Titel :
Signal and Image Processing (ICSIP), 2010 International Conference on
Conference_Location :
Chennai
Print_ISBN :
978-1-4244-8595-6
DOI :
10.1109/ICSIP.2010.5697442