Implementing Feature Extraction on Number Plates and Facial Features

Abstract

In this paper, we suggest a feature point extractor naming SURF Speeded Up Robust Features. It is mostly used for 3D images reconstruction, object identification & recognition. It performs in detecting adequate feature points in required region of interest thus making it much faster. This is performed on the basis of image convolutions and simplifying these convolutions to the necessary key points. The paper performed experiments on the various image set containing facial features and number plates. Also, it helps in performing matching by calculating SURF feature points of the current image to available image of number plate, face, eyes, nose and mouth. Both show SURF’s strong performance.