Abstract:In view of video frequency frame data in time axis which has Gaussian distribution characteristic,a clustering algorithm based on the measure of Gauss kernel function similarity and an Improved Particle Swarm Optimization(IPSO) is presented.The proposed algorithm can realize accuracy clustering by Gauss kernel function similarity measure,and speed up the clustering process by the IPSO.Experimental results show that the proposed algorithm has greater searching capability and clustering accuracy,which is superior to the C-Mean in analysis of video frequency frame data clustering.

Gauss Kernel Function Clustering Algorithm Based on Particle Swarm Optimization