Object identification using Interleaved feature extraction model

Image mining spanning the system of image applications has proved to be of good use in most industrial image processing applications. It supports a large field of applications like medical diagnosis, agriculture, industrial work, space research, and the educational field. Image Mining involves extracting information as well as image detection and extracting the image segment. It is often seen that these steps are considered in isolation leading to completely independent flow of process. This paper overlaps the steps leading of feature extraction and object recognition to provide better results of object identification. The increase in best features’ percentage is a novel idea and yields good results of identification. ORB being very fast is used in the first pass and when object is not identified for confirmation we use the SURF to confirm the results. The process of identification can easily be implemented to more than one image in a directory and hence provides better and good recognition. This bulk processing can further be extended to repository of images in the social media hence increasing the scope of the research.