Design And Implementation Of Face Feature Point Extraction And Matching Software

Feature points extraction and matching is one of the key steps in the face recognition system, the effect of which has a direct impact on the performance of the whole face recognition system.Therefore, it is very important to study it.This paper studies the feature points extraction and matching based on SIFT algorithm, and on this basis,has realized the design and implementation of facial feature points extraction and matching software.In this paper, the software system is realized by the pre-processing module, face detection module, feature point extraction and matching module.Firstly, the pre-processing module makes the face cut and alignment in the picture library.Secondly, based on the Adaboost algorithm, the face detection module is realized.Finally, in feature points extraction and matching module, based on AAM algorithm and sift algorithm have achieved facial feature points calibration and facial feature points extraction, feature points by calculating the similarity of feature point matching, so as to complete the design and implementation of the software system.The software combines the AAM algorithm calibration face and SIFT feature extraction, obtain more key feature points in face images.In this paper, we analyzes the merits and deficiencies of SIFT and propose new strategies for feature extraction and image matching, which leads to a new framework that overcomes limitations of SIFT in solving the face recognition problem.This software system provides a research platform for the further research of SIFT algorithm, AAM algorithm and Adaboost algorithm.