Byeong Gi Lee. A New Algorithm to Compute the Discrete Cosine Transform. IEEE Trans. On Acoustics, Speech, and Signal Processing, ASSP-32:1243--1245, 1984.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....the input so that the final output can be given as: 31 16 1 2 k k i Z k i Cos S p i = 0 to 31 Where Z is given by: 31 . 24 ) 24 ( 23 . 0 ) 8 ( k k S Z k The above equation is a type III DCT and can be solved efficiently by using Lee s algorithm [7]. A circular buffer has been used to handle the time domain sample values. This helped us in reducing the memory requirements as well as the complexity involved in shifting the contents. Circular buffer operations were combined with windowing, which involves point to point multiplication, to ....

Byeong Gi Lee, "A New Algorithm to Compute the Discrete Cosine Transform," IEEE Transactions on ASSP, Vol. 32, No. 6, pp. 1243-1245, December 1984.

....vertical frequencies are represented by higher row numbers and the higher horizontal frequencies are represented by higher column numbers. 1The complexity of the DCT can be simplified tremendously. For example, in the one dimensional 8 element DCT, the multiplications can be reduced to 12, see [5]; and for the two dimensional 8 by 8 case, with scaled DCT s some multiplications can be combined with the quantization process, and the number of multiplications purely for the inverse DCT is reduced to 61, see [6] Original Effective frequency Signal replication by a DCT 0 N time 0 N 2N 3N ....

Byeong G. Lee, "A New Algorithm to Compute the Discrete Cosine Transform," IEEE Trans. Acoust., Speech, Signal Processing, vol. ASSP-32, No. 6, pp. 1243-1245, December 1984.

....DCT on the columns and a one dimensional DCT on the rows. An explicit formula 2 for the two dimensional 8 by 8 DCT can be written in 2The complexity of the DCT can be simplified tremendously. For example, in the one dimensional 8 element DCT, the multiplications can be reduced to 12, see [6]; and for the two dimensional 8 by 8 case, with scaled DCT s which integrate the multiplications into the Original Effective frequency Signal replication by a DCT 0 N time 0 N 2N 3N time Figure 2.2: The DCT transform. The correspondence between the original windowed signal between 0 to N and ....

Byeong G. Lee, "A New Algorithm to Compute the Discrete Cosine Transform," IEEE Trans. Acoust., Speech, Signal Processing, vol. ASSP-32, No. 6, pp. 1243-1245, December 1984.

....A Delta B Delta A T , where B is the input block and A T denotes the transpose of A. This would involve 1024 multiplications for each input block. Nevertheless, various fast algorithms have been introduced in the literature for reducing the number of multiplications involved in the transform [20]. The algorithm used in the JPEG standard only needs to perform 192 products to produce one resultant block; but because of this optimization, the new code cannot be vectorized. The use of memory tables to replace multiplications or other operations groups is quite frequent too. In color space ....

Byeong Gi Lee. A new algorithm to compute the discrete cosine transform. IEEE Transactions on Acoustics, Speech, and Signal Processing, vol. ASSP-32, 6:1243--1245, Dec. 1997.

....DCT on the columns and a one dimensional DCT on the rows. An explicit formula 2 for the two dimensional 8 by 8 DCT can be written in 2 The complexity of the DCT can be simplified tremendously. For example, in the one dimensional 8 element DCT, the multiplications can be reduced to 12, see [6]; and for the two dimensional 8 by 8 case, with scaled DCT s which integrate the multiplications into the 5 0 N 0 N 2N 3N Original Signal Effective frequency replication by a DCT time time Figure 2.2: The DCT transform. The correspondence between the original windowed signal between 0 to N and ....

Byeong G. Lee, "A New Algorithm to Compute the Discrete Cosine Transform," IEEE Trans. Acoust., Speech, Signal Processing, vol. ASSP-32, No. 6, pp. 1243-1245, December 1984.

....B Delta A T , where B is the input block and A T denotes the transpose of A. This would involve 1024 multiplications for each input block. Nevertheless, several fast algorithms have been introduced in the literature aimed at reducing the number of multiplications involved in the transform [20]. The algorithm used in the JPEG standard only needs to perform 192 products to produce one resultant block; but because of this optimization, the new code cannot be vectorized. The use of memory tables to replace multiplications or other operation groups is quite frequent too. In color space ....

Byeong Gi Lee. A new algorithm to compute the discrete cosine transform. IEEE Transactions on Acoustics, Speech, and Signal Processing, vol. ASSP-32, 6:1243--1245, Dec. 1997.

....[16, Appendix A] For the row column decomposition method, the 2 d idct is calculated by performing sixteen 1 d 8 point idcts. Efficient algorithms can be based on using the fast Fourier transform (fft) as a basis for computing the dct [21, 22] or on the properties of the transform itself [23, 24]. f(x; y) 2 N N Gamma1 X u=0 N Gamma1 X v=0 C(u)C(v)F (u; v)cos (2x 1)u 2N cos (2y 1)v 2N (1a) where C(u) C(v) ae 1 p 2 for u; v = 0 1 otherwise (1b) For real time mpeg 2 decoding with the d30v processor, the software must exploit hardware parallelism ....

Byeong Gi Lee. A new algorithm to compute the discrete cosine transform. Trans. on Acoustics, Speech, and Signal Processing, 32(6):1243--1245, Dec. 1984.

No context found.

Byeong Gi Lee. A New Algorithm to Compute the Discrete Cosine Transform. IEEE Trans. On Acoustics, Speech, and Signal Processing, ASSP-32:1243--1245, 1984.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC