A Unified Approach to Data Modeling and Retrieval for a Class of Image Database Applications

摘要: Recently, there has been widespread interest in various kinds of database management systems for managing information from images. Image Retrieval problem is concerned with retrieving images that are relevant to users’ requests a large collection images, referred as the image database. Since application areas very diverse, seems be no consensus what an system really is. Consequently, characteristics existing have essentially evolved domain specific considerations [20]. In response this situation, we introduced unified framework retrieval databases [17]. Our approach based on premise it possible develop data model and associated can address needs class applications. For applications, perspective end users, processing two orthogonal issues distinction contributes toward domain-independence. paper, analyze approaches modeling establish taxonomy which these systematically studied understood. Then investigate applications view point their requirements both attributes generic types. To support types, proposed model/framework AIR. AIR employs multiple logical representations. The representations viewed abstractions physical at levels. They stored persistent We then discuss how developed framework. Development our implementation briefly described. Finally, identify several research solutions some them indicated.