Implementation of A Low-Cost Embedded Multi-Camera System for Leaf Plant Monitoring in the Greenhouse
Aryuanto Soetedjo 1, Evy Hendriarianti 2, Muhammad Suriansyah 1, M Rifki Abdilah 1, M Syahriel M 1, and Mohamad Khafil HM 1

1) Department. of Electrical Engineering, National Institute of Technology (ITN) Malang, Jalan Raya Karanglo KM 2, Malang, Indonesia
2) Department. of Environmental Engineering, National Institute of Technology (ITN) Malang, Jalan Bendungan Sigura-gura No. 2, Malang, Indonesia

Abstract

The leaf is an essential part of a plant that can indicate plant growth and phenotyping. Real-time leaf plant monitoring provides significant benefits in this field. This paper presents an implementation of the multi-camera system for leaf monitoring in a greenhouse. A low-cost multi-camera system consisting of the visible (RGB) camera, infrared camera (NoIR: No infrared filer), and thermal camera are employed as imaging devices. Each camera is interfaced with an embedded Raspberry Pi module, which then processes and sends the image data to the Google Drive cloud. The cameras and Raspberry Pi modules are arranged and installed on a portable platform that can easily be placed in a greenhouse. The camera can be positioned accordingly to get the optimal field of view. The multi-camera system has been tested on the greenhouse for monitoring the Ramie plant continuously for twenty-one days, where the data are sent to the cloud every five minutes. The experiment showed that the image data are sent to the cloud storage with a success rate of 99.7%, 95%, and 99.8% for the RGB, NoIR, and thermal cameras, respectively. The results showed that the developed multi-camera system is reliable for leaf plant monitoring in the greenhouse..

Keywords: Multi-camera,leaf monitoring, Raspberry Pi, IoT

Topic: Computer engineering, vision and application