High Dimensional Indexing in Image Databases With Adaptive Cluster Distance Bounding

The authors consider approaches for exact similarity search in a high dimensional space of correlated features representing image datasets, based on principles of clustering and vector quantization. They develop an adaptive cluster distance bound based on separating hyper-planes that complements their index in selectively retrieving clusters that contain data entries closest to the query. This bound enables efficient spatial filtering, with a relatively small pre-processing storage overhead and is applicable to Euclidean and Mahalanobis similarity measures.