Key points are not available for this paper at this time.
The availability of easy-to-use, low-cost, and highly scalable tools makes it possible to achieve rapid and widespread adoption of precision agriculture. In this paper we outline the development of a smart camera system to detect drought stress in corn and soybean crops. The system is comprised of a Raspberry Pi Zero W, Raspberry Pi Camera, WittyPi mini, a cooling and solar power system, temperature sensors both inside and outside of the box, and infrared canopy temperature and light sensors. The system was built to collect data in a configurable time frame and has an embedded machine-learning (ML) processing system. The camera was configured using an Internet of Things (IoT) platform to manage the device and send images to the Cloud. One of the challenges for this system was to effectively implement machine learning models on this limited-resource embedded platform. We achieved an accuracy of 74% with the embedded machine learning algorithm when classifying water stress in soybeans.
Ramos-Giraldo et al. (Sun,) studied this question.