Hybrid Feature Based Object Mining And Tagging

Image Tagging are important as far as image search engines/databases are concerned viz. Flicker, Picasa, Facebook...etc. Image Tagging is a difficult and highly relevant machine learning task. Image tagging with algorithms based on `Nearest neighbor classification' have achieved considerable attention on the implementation point of view but at the cost of increasing computational complexity both during training and testing. In the existing approaches used single object based tagging. In this research paper we are going to discuss different research related to object mining and tagging. As far as there are shape, color and texture feature are impotent to describe object. The proposed system firstly use KNN for tagging different object features for training. Using color moment, shape and gray level co-occurrence matrix (GLCM) as a texture feature. After that system will use adaboost classifier for classification of objects and final image represented by different object tags.