Smart Door Implementation Using Jetson Nano-Based OpenCV and Deep Learning

초록

This paper implements a smart door using face recognition based on Jetson Nano. Existing digital door locks are locked and unlocked using a password. As security issues such as the recent increase in single-person households increase, the need to use new technologies for home security, such as implementing a smart city, is emerging. Face recognition-based home security systems have become an effective approach due to the recent development of deep learning technology. Jetson Nano is a small artificial intelligence computer that is an intelligent embedded system with great utility for smart door development. The proposed smart door can provide safety and security that it recognizes faces and controls the door using these latest technologies. The proposed smart door uses two main technologies to provide home security. One is to recognize faces using OpenCV-based deep learning technology, and the other is to directly control the door lock through Jetson Nano's GPIO interface. In the proposed system, home security can be provided by connecting and controlling the face recognition system and the door lock. The experimental results show that the proposed system can effectively control the door by recognizing a real face from a camera image. © 2021, Korean Institute of Communications and Information Sciences. All rights reserved.

키워드