Face Recognition under Illumination based on Optimized Neural Network

Face recognition is a significant area of pattern and computer vision research. Illumination in face obvious yet challenging task matching. Recent researchers introduced machine learning algorithms to solve illumination problems both indoor outdoor scenarios. The major challenge the lack classification accuracy. Thus, novel Optimized Neural Network Algorithm (ONNA) used aforementioned drawback. First, we propose Weight Transfer Ideal Filter (WTIF) which employed for pre-processing remove dark spots shadows an image by normalizing low frequency high illumination. Secondly, Robust Principal Component Analysis (RPCA) perform efficient extraction features based on representation. These are given as input ONNA classifies under Thus achieve various conditions. Our approach analyzed compared with existing approaches such Support Vector Machine (SVM) Random Forest (RF). better terms accuracy error rate.