Crop Assessment Technique Using a Real-Time Hardware-Based Image Processing System to Support on-the-spot Decision Management for Agricultural Applications

Abstract

Despite the advances in modern technology, one of the most challenging tasks remains the identification of stressed and/or diseased crops at the field scale. A wide range of precision agriculture (PA) technologies with the integration of remote sensors, Global Positioning System (GPS), and Geographic Information System (GIS), are continuously serving the agriculture industry. Though PA technologies are still evolving, there are some limitations to efficiently applying PA solutions at the field scale, including high computational cost, high complexity, low image resolution, and low GPS accuracy. While Field Programmable Gate Array (FPGA)-based flexible hardware solutions can be a cost-effective and simple alternative to the current multiprocessor systems, it is not widely used in PA. On the other hand, Real-time Kinematic GPS (RTK-GPS) provides the opportunity to achieve cm level accuracy for geographical data collection purposes. Therefore, this study is based on the integration of an FPGA-based real-time image processing system, along with highly accurate RTK-GPS data to develop a real-time crop monitoring system. The result is a decision support system that delivers a heat map to the end-users based on the intensity of the detected parameter from the field of interest to assist in site-specific farm management solutions. Through this study, the developed system proved its potential to resolve some of the current PA limitations to provide on-the-spot decisions by combining FPGA-based image acquisition and processing along with high positional accuracy.