Journal ID : TRKU-05-04-2020-10649
[This article belongs to Volume - 62, Issue - 03]
Total View : 212

Title : FACE RECOGNITION USING STATISTICAL FEATURE EXTRACTION AND NEURAL NETWORK

Abstract :

In this study, a human face recognition technique based on statistical features using a neural network technique is presented. In the pre-processing stage image edges have been detected. Subsequently, a new technique for two-dimension gray image to one-dimension vector is proposed. Then, seven features have been extracted depending on statistical analysis. This work describes is based on four statistical characteristics (mean, standard deviation, skewness, kurtosis) for feature extraction. These features can address image capture problems because it is main tasks are not affected by the rotation, zoom, and transfer of images taken from before the control cameras. After that, the image details have been extracted using wavelet transformations. Elman Neural Network (ENN) is used in this study for face identification. Finally, the proposed study has been implemented using MATLAB R2013a and Microsoft Excel database to preserve the information of the required people, this can be achieved by utilizing the principle of distance between facial points

Full article