High Impact Factor : 4.396 icon | Submit Manuscript Online icon |

A Reconfigurable and Scalable Architecture for Discrete Cosine Transform

Author(s):

Maitra S Aldi , VTU PG CENTRE KALABURAGI; Dr. Baswaraj Gadgay; Zameer Ahamad B.

Keywords:

Discrete cosine transforms, Approximation of DCT, Scalable DCT, Reconfigurable DCT, Image compression

Abstract

A design for a general and reconfigurable structural design for approximation of DCT (Discrete cosine Transform) is proposed. Approximation is needed to reduce complexity of calculations in DCT without altering its performance. An algorithm is presented in this project for obtaining approximate DCT coefficients of length N from two N/2 lengths of DCTs with N number of additions for preprocessing of input. An algorithm proposed here is highly scalable for software implementation and for hardware implementation as well and involves lesser computational complexity to achieve several power of 2 DCT. The presented parallel architecture to compute approximation of DCT is fully reconfigurable; it is configured to compute 32-coefficient points DCT along with two parallel 16- coefficient points DCTs or four parallel 8- coefficient points DCTs. The scalable design is used in compression of different size of images.

Other Details

Paper ID: IJSRDV4I50801
Published in: Volume : 4, Issue : 5
Publication Date: 01/08/2016
Page(s): 1248-1252

Article Preview

Download Article