Single Sample Face Recognition Based on QR Decomposition and 2DLDA

This paper proposes a new single sample face recognition algorithm based on matrix QR decomposition and 2DLDA. The algorithm uses QR decomposition of a matrix on the single sample, extracts the effective information to constitute a virtual image, combines the original training images and generates a new training sample set, then the application of 2DLDA algorithm for feature extraction and recognition. The algorithm is tested on ORL face database, experimental results show that this algorithm is not only superior to the traditional algorithms on recognition rate, such as PCA, SPCA, (PC) ²A, E(PC)²A algorithms, but also has good robustness for factors such as illumination, expression etc.