Mining images for image annotation using SURF detection technique

Image annotation is a method to title a particular image. The image annotation scheme would create a scenario in which a set of images are provided as training set. The query image is then identified and properly annotated, by matching the features of the query image and the existing training set images and correspondingly labels are provided. This work aims at working on the annotation of different types of images using training set images, the approach is to calculate the feature set of the training set images and saving them into an array. The query images feature set is also calculated and compared with that of the existing ones, the one with more matching features would be highlighted and the corresponding label of that particular image is provided to the query image. The analysis of this particular technique is performed with large number of test image sets for better accuracy.