This is a demonstration of the PowerPoint slides and the program codes for the course project for EEL6825 Pattern Recognition at the University of Florida.
For the purpose of this project, we shall delve deeper into facial detection and identification, by implementing an algorithm to help the computer accurately identify a person in real time using the web camera and a database of training photographs, and distinguishing between individuals. Access control is regulated using the algorithm developed. The accuracy of the algorithm is the main evaluation metric of the system, and the highest accuracy achieved is 93.94%, and the average accuracy (depending on the lighting conditions, environment, webcam quality, etc.) is 89.27%. Python 3.7.1 is used from Anaconda 2018.12's Spyder. The OpenCV library, along with its Haar Cascade classifier, is the primal face detection classifier, while the Local Binary Patterns Histograms (LBPH) Face Recognizer is the main facial recognition classifier being used for the algorithm.