This paper examines the problem of efficient indexing of large image databases using the theory of optimal keys. The methods based on optimal keys are compared empirically with a texture classification method and template matching for benchmarking purposes in the Leiden WWW color image database and the 19th century portrait database. The different indexing methods are compared and evaluated in the problem space of finding copies of corrupted images. Real world noise is present in the form of print-scanner noise and general image degradation.
CITATION STYLE
Lew, M. S., Huijsmans, D. P., & Denteneer, D. (1997). Optimal keys for image database indexing. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1311, pp. 148–155). Springer Verlag. https://doi.org/10.1007/3-540-63508-4_117
Mendeley helps you to discover research relevant for your work.