Efficient Algorithm for Reducing the Computation Redundancy of LBG Algorithm
Chao Pang
Abstract
Chao Pang
Abstract
A fast codebook generation algorithm named fast local clustering algorithm is presented in this paper by the equivalence of norm and the property that data cluster locally. The experiment by the typical test picture Lena and Barb shows that the algorithm is 3.74~9.59 times faster than LBG, and PSNR of it is only 0.25~0.43dB smaller than PSNR of LBG. It is the best algorithm among the six type algorithms introduced in this paper.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
A fast codebook generation algorithm named fast local clustering algorithm is presented in this paper by the equivalence of norm and the property that data cluster locally. The experiment by the typical test picture Lena and Barb shows that the algorithm is 3.74~9.59 times faster than LBG, and PSNR of it is only 0.25~0.43dB smaller than PSNR of LBG. It is the best algorithm among the six type algorithms introduced in this paper.
Key concepts: Algorithm, Codebook, Linde–Buzo–Gray algorithm, Computation, Cluster analysis, k-medoids, Redundancy (engineering), Computer science