[M-GI31] Earth and planetary informatics with huge data management

[MGI31-P07] Programming Techniques for visual IoT on Raspberry Pi (6)

Toshiki Aoki¹, *Takamichi Mizuhara¹, Yuki MURAKAMI², Kazutaka Kikuta², Kazuya Muranaga⁴, Kazunori Yamamoto², Tomohiro Kawanabe², Ken T. Murata², Keiichiro Fukazawa³ (1.CLEALINK TECHNOLOGY CO., LTD., 2.National Institute of Information and Communications Technology, 3.Kyoto University, 4.Systems Engineering Consultants Co.,LTD. )

Keywords:HpVT, Visual IoT, Raspberry Pi

The recent emergence of low-cost single board computers such as Raspberry Pi and Arduino, which are used as sensing devices with connectivity to Internet gateways and cloud services, has led to Internet of Things (IoT) architectures in the real world. Visual IoT is one of the classes of IoT that poses crucial end-to end challenges due to the need for sensing and processing of visual data. One of the significant roles of the visual IoT is information extraction from images by using image recognition techniques.
In this paper, we present a programming environment of visual IoT on Raspberry Pi (RPi) and develop a couple of applications. So far, our system of video streaming requires a decoder on RPi working as a receiver. To enhance usability, we develop an WebRTC environment so that users receive the streaming video on their PC terminal or smartphone in real-time. In addition, we propose a Web application to handle PTZ (pan-tilt-zoom) operation of the IP network camera on the WebRTC streaming video. For example, users make a choice of a position on the video by mouse, then centralization and enhancement of the video is carried out.