Inverse discrete cosine transform architecture exploiting sparseness and symmetry properties

In this paper, a novel architecture for 2D IDCT is proposed, based on the sparseness property of 2D DCT coefficient matrix and the even and odd symmetry property of the basis vectors of the 1D DCT transform. The proposed architecture performs 2D IDCT directly on the 2D DCT coefficient matrix to avoid timing and area overheads of the transposition. We derive a recursion equation from the definition of the 2D IDCT algorithm and use it to design an efficient 2D IDCT architecture. The proposed architecture consists of highly regular, parallel and pipelined elements which are suitable for VLSI implementation. It is shown that the proposed architecture can achieve a high throughput rate and a low hardware complexity, when compared with other DCT-based IDCT architectures. Another important aspect is that the proposed architecture can provide an efficient way to control the trade-off between visual quality of the reconstructed image and computational complexity.