Abstract: In Artificial Neural Network (ANN) experiments, the system of face recognition consists of image preprocessing, image segmentation, detection and feature extraction, localization and normalization and ANN. First, the process of preprocessing for human face is an improvement of the image data and image segmentation is the first step of image analysis that seeks to simplify an image to its basic component elements or objects. Then, the detection and feature extraction is the process of local feature and global feature for human face to extract features in localization and normalization. Finally, a back-propagation ANN is used and trained to recognize the actual human face images.