Abstract: The recent growth in the volume of image data being generated and used for a variety of applications have insisted the development of image databases. The richness of content and subjective interpretations of image has rendered text based queries inadequate. Content Based Image Retrieval ( CBIR) is a new but widely adopted method for finding images from vast and unannotated image databases. The correctness of retrieval for CBIR depends on efficient and effective indexing and searching schemes. Subjective queries and retrieval demands enormous computation time due to large data sizes of images coupled with large and complex indices required for search. Networks of Workstations (NOW) are a cost effective way of providing the much needed computational power in such applications.This study presents a distributed scheme for the classification and retrieval of images in an image database using NOW system. It uses an initial classification and a heuristic for determining the average feature vectors and distance in that image classes. The results of classification, retrieval, speedup obtained and the correctness of the retrieval are presented. The results indicate the viability and effectiveness of the proposed scheme.
S. Baulkani and L. Ganesan , 2007. Distributed Algorithms for Image Data Base Classification and Retrieval Using Perceptual Features. International Journal of Soft Computing, 2: 339-345.