About this document
Content-Based Image Retrieval Overview by watson191 is a document available to read on EtoBox.
Content-based image retrieval (CBIR) analyzes image contents like color, shape, and texture rather than metadata to search image databases. Early CBIR systems searched based on color, texture, and shape, but now incorporate human feedback to better understand user intent. CBIR compares images using distance measures in various dimensions and remains an open problem, though techniques like color histograms, texture modeling, and shape descriptors have been developed.
- Author
- watson191
- Language
- EN