The number of digital images increases rapidly, hence to categorize these images and to retrieve the relevant web images are a difficult process. This paper uses annotation mechanism to overcome the problem on images. Here two mechanisms are followed they are manually annotating the images by using human interface and the annotated images are stored in repository. Automatic annotation they are obtained by performing feature extraction and clustering algorithm. For feature extraction SIFT algorithm is used. Last mechanism is learning annotations by clustering.

No abstract