Development of Wide-Angle Fisheye Vision System for Smart Home Security: An Initial Study

In this paper, we report the initial study of developing a wide-angle fisheye vision system for security that supports a smart home system. The developed system consists of 1) a customizable end-node, 2) OpenCV as a general infrastructure, 3) and an Android app equipped in a smartphone as a user interface (namely a Smart Security System app) for online monitoring purposes. The Smart Security System app covers two features: user register-login and recent monitoring video. The end-node part includes a PIR sensor, Raspberry Pi, and Fisheye wide-angle camera module, all packed based on the consumer product. Since the limitation of Raspberry Pi webcams or smartphone cameras led to monitoring blind spots, the research using fisheye cameras demonstrated a wider angle and presented a practical solution for minimizing blind spots in smart home security systems. The discussion of this paper focuses on a) the brief description of the proposed system, b) the comparison of the Fisheye wide-angle camera vs. the HQ camera, c) and the usability testing of the developed Android app. In addition, laboratory tests also have been observed. Since the proposed system is an initial study, we conducted a demo of the camera's angle coverage in a lab scale test. We used overall system usefulness, interface quality, and satisfaction for the app test.