Fast image retrieval with feature levels

Digital images have many applications in different fields like medical imaging and diagnostics, weather forecasting, space research, military etc. The number of images available and their wide variety increases with the ease of acquiring, storing and sharing digital images due to the advances in technology. As a result, the significance of image retrieval algorithms and systems has been considerably increased. Many researches on content-based image retrieval (CBIR) are being carried out. In this paper, a fast image retrieval algorithm called feature levels is proposed. Feature levels algorithm works with the classification of image features to different categories or levels, feature extraction in terms of levels and feature similarity comparison of the query image with database images. The system retrieves images from the associated database. The database is re-written after each level according to Database Revision (DR) algorithm.