High-dimensional query technology is an important multimedia database application. The traditional inquiry are often faced with the difficulty of " curse of dimensionality ".Firstly, the high dimensional space based on the difficulty of "curse of dimensionality", a method of using the one-dimensional mapping, and in a multi-dimensional map of the thinking for the improvement of one-dimensional map. In the following is the important improvement:1. A new one-dimensional mapping approach based on Principal Component and Distance. This method is based on the iDistance method and using principal component combined with distance named PCA-Distance, PCA transformation through effective one-dimensional map, so as to achieve fast query results. By clustering the data set, the PCA transformation index, using principal component combined with the distance for the nearest neighbor query. Experimental results show that using this method can effectively reduce the amount of data access and improve the nearest neighbor query performance.2. Multi-reference-points distance mapping methods. This method is mainly related to vector quantization in the codeword search process. The results show that: using Multi-reference-points method can effectively reduce the codeword search in the amount of data access, accelerated codeword search process. |