Textural Features and Relevance Feedback for Image Retrieval

This paper focuses on the retrieval of complex images based on their textural content. We use GMRF for texture discrimination and a region-growing algorithm for texture segmentation. Relevance feedback is introduced to improve retrieval accuracy. Keywords Textures, relevance feedback, image features, Gaussian Markovian Random Fields.

• Type: Conference
• Year: 1998
• Where: VDB
• Authors: Eugenio Di Sciascio, Giacomo Piscitelli, Augusto Celentano
• Keywords: Database, Markovian Random Fields, Relevance Feedback, Texture Discrimination, VDB 1998