Abstract :
Retrieval of images plays a major role in different domains including medical diagnosis, biometrics, industry inspection, geographical information satellite systems, web searching and historical research and so on. When size of the database increases continuously, the applications involving images face new challenges and crucial problems such as storage management, indexing, knowledge management and retrieval presentation. We need an efficient retrieval mechanism to retrieve images from the multimedia database. CBIR - Content-based image retrieval is an image retrieval technique used to retrieve images efficiently by using low level image features texture, shape and color. In CBIR system, an image query is characterised by primitive, logical features and abstract attributes. Queries with primitive features are directly derivable from the images but logical and abstract features require logical inference and objects semantic reasoning. This survey paper focuses on different features descriptors for image retrieval and analysis of various retrieval operators like LBP-Local Binary patterns, LTP-Local Ternary patterns, LDP-Local Derivative Patterns and LTrP-Local Tetra patterns using high level features to improve the performance and accuracy in CBIR system.
Keywords :
content-based retrieval; feature extraction; image retrieval; CBIR system; LBP; LDP; LTP; LTrP; content based image retrieval; features descriptors; local binary patterns; local derivative patterns; local ternary patterns; local tetra patterns; retrieval operators; Feature extraction; Histograms; Image color analysis; Image retrieval; Shape; Visualization; Content based image retrieval; Local Binary patterns; Local Derivative Patterns; Local Ternary patterns; Local Tetra patterns;