Resumen

Recently, very large collections of images and videos have grown rapidly. In parallel with this growth, content-based retrieval and querying the indexed collections are required to access visual information. Three main components of the visual information are color, region and contour. In this thesis, Content-Based Logo Image Retrieval Using Visual Mpeg-7 Descriptors; is presented a tool that combines three visual descriptors of MPEG-7 and measures similarity of images by applying a distance function calculated by de software XM (eXperimental Model MPEG-7) to the efficient retrieval of logo images . An XML description is used for storing the three descriptors measures. For the image retrieval process, the XML descriptions are compared to the query description and then a new XML description will be generated this will store the distances values between the repository images and the query. The farthest image to the query is used for the results normalization process since the descriptions are arranged in the most distant from the nearest of the query, another relevant aspects as the weighting of the descriptors depend on the search configurations. Also it is shown through experimental results and analysis of retrieval effectiveness of querying that the content-based retrieval system is effective in terms of retrieval and scalability.