A Survey on Automatic Image Annotation and Retrieval

The Image Annotation process are required to use automated where the strong tagging is required to keep annotate image for making it efficient to provide better results while querying those image annotated database. The process of image annotation should be proceeding while creating the file and to make it strongly labelled. It is a process of machine learning where low level features of images are extracted, clustered and mapped to the semantic. This can be based on training set of data. Automatic image annotation technique can be based on various things either it can observe the images various ways either in texture bases, colour intensity basis or faces included or involved into the images. Next the features are grouped into the cluster and then annotation is done based the category. In this paper we study & describe the how the different automation techniques are working in order to annotate the datasets and then how they are useful in order to query process the data and to release the query work load and to reduce the computation process time. We have compared different annotation techniques on various parameter like segmentation, feature extraction, clustering etc. Furthermore number of models has been observed and reviewed in order to face the challenges being found in those and got rectified in proposed approach. Keywords-Automatic image annotation, image Tagging, Image Query, image retrieval, annotation models.