Spring growth stage detection in Italian ryegrass field using a ground-based camera system

Abstract

Information on the spring growth status of winter forage crops is crucial for evaluating productivity and nutrient management. This study aimed to determine the spring quick growth stage (QGS) of Italian ryegrass using a ground-based camera system. The camera system, installed in two Italian ryegrass fields at the farm of Hiroshima University, captured images automatically three times per day in red, green and blue channels over the growing season in 2012–13. Four presumed color intensities/indices were fitted using a logistic model to construct smoothed time-series data. Among the color intensities/indices, excess green was suggested to be the best parameter for monitoring seasonal changes. The root mean squared error of the estimated phenology dates against plant height was 7.7 days for the start-QGS and 2.8 days for the end-QGS. These results from a single year should be broadened to examine other methodologies for image processing and extended to multi-year data.