Face detection and recognition in real-time photos with haar cascade and local binary pattern histogram for automatic door locking system
Ian Haikal A, Paqih Raihan, Wiwit Rahayu, Dea Febrianty, Nailil Farihah, Winda Azizah, Iwan Cony Setiadi, Yonatan Davidson Gultom · 2021
Home door security systems using facial recognition based on image processing have been widely developed. Face detection in this system uses the Haar-Cascade Method. The system was tested using 4 types of tests, namely the accuracy test, the distance test, the facial expression test, and the lighting conditions test. The results show that this door design system has a total average accuracy level of 97.2%, with an optimal distance of 1.5 meters, and the condition of the lights must be on. Meanwhile, for the facial expression variation test, the system can distinguish well except when the face is tilted left or right.