3D face recognition using Hadoop
G. Geetha, Mohammad Baradaran Safa, C. Fancy, K. Chittal · 2017
Face Recognition is one of the biometric technique to vestige the given faces. We present a 3D face recognition method using Hadoop to recognize 3D faces under varying expressions, lighting and different poses to overcome the challenges of the 2D face recognition. In this paper, a threshold facial region of an image is detected and preprocessing is done based through the image excellence. If the selected face is frontal face with good lighting, extract the prerequisite features and do the necessary comparison steps to recognize the faces. In case, if the selected face is in bad lighting, then perform histogram equalization and normalization to increase the contrast. Different Poses and expressions are the very challenging zones which require surplus pre-processing to improve the performance of the face recognition system. Hence an enhanced normalization method called 3D Morphable Model are used as a pre- processing technique to create a frontal view from a non-frontal view and also merge images with different views in to a single frontal view. Next to diminish the number of features used for recognition process; we emulate the linear discriminant analysis method for further classification. Eventually, we used an open-source Hadoop Image Processing Interface (HIPI) to act as an interface for MapReduce technology for recognition.