Abstract: This project entitled "Unknown Person Identification and Alert system for Blind people using face Recognition " is developed using Python programming language. Security is a very important term because most crimes are taking place in cities rather than in rural areas. So it is compulsory to appoint a watchman or security guards, which is not affordable for everyone. IOT-based security and technological substitutions have easy but much costlier maintenance requirements. The primary goal of this project is to develop a software application that will help identify intruders using Opencv. Computer vision and OpenCV have mitigated this problem through the proposed system. The system can detect and recognize a person based on the data previously provided to it. The proposed system is capable of face detection and reorganization of a person to identify whether the person in the frame is a known person or an intruder. Haar-cascade models are used for detecting human face. This portion of a face is then given for recognition through a model trained on LBPH. The model returns the probability of recognition in the frame for each person. Furthermore, the system can trigger the alarm when it comes to finding any intruder concerning human beings. This can be integrated into any kind of video monitoring like safety to banks, burglar movement monitoring applications, etc. The system will definitely require some kind of initial training to learn what is normal and then an intruder. The development of this system is mainly to make it user-friendly and by that, reduce the amount of manual monitoring work possible.

Keywords: Face Recognition, Computer vision, Haar-cascade, Local Binary Pattern Histogram (LBPH)

| DOI: 10.17148/IJIREEICE.2025.13313