Leaf Temperature Monitoring System Using Low-Cost Thermal Camera and IoT Technology

Leaf temperature provides valuable information about the water status of the plant. A leaf temperature monitoring system will offer potential benefits in precision agriculture. This paper presents an implementation of low-cost cameras, namely MLX90640 and AMG8833, for monitoring the leaf temperature. The low-cost camera is connected to a Raspberry Pi Zero W, a tiny single-board computer that performs the data processing and interfacing to the ThingSpeak IoT platform. The experimental results show that even though the AMG8833 has a low resolution of 8×8 pixels, it can be used to measure the leaf temperature, similar to the MLX90640, whose resolution is 32×24 pixels. The AMG8833 achieves a refresh rate of 26.57 fps, very fast than the MLX90640, which achieves 2.10 fps. Moreover, three thresholding algorithms, Otsu, Yen, and Mean thresholding work properly to separate the leaf from the background. Thus the leaf and background temperature can be defined easily.