A clustering algorithm based on best-first search strategy

This paper studies and analyses the problems and deficiencies that are faced by clustering algorithm which is based on hierarchical scene tree, and proposes a clustering algorithm that is based on the best-first search strategy. The algorithm reduces the false-matching rate, and also alleviates the problem of network congestion to some extent, by designing a novel cost function. Both theoretical analysis and experiment reveal that the method of clustering based on best-first search strategy has better data processing capability, and it is more suitable for the application of large-scale virtual environments comparing to the original algorithm based on hierarchy scene tree.