Image processing plays a pretty important role. Much research has been carried out on image objects, where information from images can contribute to and benefit education, technological innovation, and information management. The face is a marker of recognizing someone. Facial recognition systems have difficulty with different facial orientations, lighting, haircuts, mustaches or beards, glasses, permanent blemishes or scars, and differences in conditions such as the person turning slightly, looking down, or looking up. The research aims to carry out digital image processing, specifically the process of scaling, grayscale, edge detection, and thresholding the image of a person's face as input. This research is crucial because image processing, especially facial recognition, is essential in health, education, economics, and security. The research uses the Artificial Neural Network method. Artificial Neural Networks are information processing systems designed to imitate how the human brain works in solving a problem by learning by changing the synapse weights. Research results in facial recognition using backpropagation show high accuracy and a fast average time. The suitability level of the recognition results depends on the combination of parameter values used in the learning process. The greater the epoch (repetition), the greater the learning rate; the smaller the error, the higher the known level.
Copyrights © 2024