Feature Extraction Using Histogram of Oriented Gradient and Hu Invariant Moment for Face Recognition

摘要

Face recognition was one of many popular fields in Image Processing and Computer Vision. Face recognition had many problems like the pose variation, Illumination, Image Quality, etc. Many methods were developed to find the solution to this problem. Some method uses Approach that uses the whole face as features, while other method uses Features-Based Approach that uses Local Feature like Eyes, Nose, and Mouth as features. The proposed method to combine two methods: HOG and Hu Invariant moment. Two methods as feature extraction has been implemented with tested on 3 database: Markus's, ORL/AT&T and our database. This database was tested with three scenarios testing. Finally, the proposed method result on Recognition Rate average is 79.82 %, then the best Recognition Rate result is 97.22 %.