An Automatic Image Annotation Technique Based on Coverage Ratio of Tags

In this paper we propose an automatic image annotation technique. The proposed technique is based on coverage ratio of tags by employing both image content and metadata. The images in a reference image set are employed to automatically annotate a given image according to the coverage ratio of the tags in the reference image. Tags and content descriptors as well as coverage ratio of the tags are generated for the entire images in the reference image set. The color and texture content descriptors are employed to retrieve similar images for an un-annotated image from reference image data set and the tags in the metadata of the retrieved images are used to annotate the un-annotated image. Simulation results indicate that the proposed technique outperforms another automatic annotation technique that uses similar content descriptors both in average precision and average recall.