An Approach To Automatically Generate Digital Library Image Metadata For Semantic And Content- Based Retrieval

Metadata represents textual information attached to an image or resource to aid identification and retrieval of that resource. In this paper it is revealed an approach to automate the creation of digital library image metadata embedding semantic and content features. These features will make more precise image content indexing and will allow fast retrieval of images in digital libraries, based on similarities with a given image. The proposed system will have the capacity to automatically generate Digital Library images metadata embedding visual features extracted after image processing. This system will read features arrays, previously created by image processing algorithms and saved them in XML documents that describes the image metadata. Further, using the XSLT(eXtensible Stylesheet Language for Transformations) language, the developed system will automatically generate digital library image metadata allowing the Content-based Image Retrieval in Service Providers.