作者: V. Ramasubramanian , K.K. Paliwal
关键词: Weighted Voronoi diagram 、 Centroidal Voronoi tessellation 、 Computational geometry 、 Voronoi diagram 、 Algorithm 、 Lloyd's algorithm 、 Proximity search 、 Bowyer–Watson algorithm 、 Mathematics 、 Nearest neighbor search 、 Signal processing 、 Electrical and Electronic Engineering
摘要: Abstract In this paper we consider fast nearest-neighbor search techniques based on the projections of Voronoi regions. The diagram a given set points provides an implicit geometric interpretation and serves as important basis for several proximity algorithms in computational geometry developing structure-based vector quantization techniques. provide approximate characterization regions with respect to their locus property localizing small subset codevectors. This can be viewed simplified practically viable equivalent point location using while circumventing complexity full diagram. paper, comprehensive study two projections, namely, box-search mapping table-based context encoding. We also propose effect advantage principal component axes data high degree correlation across components, reducing projections.