Design of Home Security System Using Face Recognition with Convolutional Neural Network Method

Best Machine Learning Model For Face Recognition in Home Security Application

“…If the video captures the face of people, the system crops the face image and then is processed the classification to recognize the face. In the previous research [5], the deep learning method was used for classification, which has complex method, it needed the other option for perform face classification, in this study is machine learning. This study uses a simple machine learning model to look for simple alternatives to recognizing the face.…”

“…This research is expanded from previous research [5], which has a system that must recognize people's faces who live in the house or strangers before the system unlocks the door. The dataset image of the face is needed subject to the learning process machine learning model to recognize the system.…”

“…Criminal cases have increased by 7.3% in 2022 compared to previous year in Indonesia [1], so further prevention is needed to solve this problem. Many things have been done to improve home security, from adding monitoring systems such as CCTV [2]- [4], until developing smart lock doors to monitor who enters the house [5]- [8]. All these systems are supported by face recognition technology to monitor who is in the house.…”

“…After that, to get the face feature, face segmentation is executed. That feature is used to recognize many classification method options [2]- [5], [8]- [10]. Accuracy identification face becomes most important in the security system to prevent miss classification and make the system more reliable.…”

“…A previous study [5] has designed a home security system with deep learning methods, in this case CNN is used to recognize householders and strangers. That system creates a good classifier model with an accuracy reach of 99%.…”