An image processing approach for measurement of chili plant height and width under field conditions

Abstract Plant height and width is an essential phenotypic parameter that can be used not only as an indicator of overall plant growth but also used to estimate the advanced parameters such as the design of agricultural machines, estimation of yield, and site-specific applications. Presently, chili plant height and width are mostly measured manually, which is laborious and time-consuming process. The goal of this study was to develop and evaluate a real-time phenotyping system using an image processing approach to measure chili plant height and width under field conditions. The image processing algorithm was developed and compiled in the open-source computer vision library (OpenCV) and Python language using PyCharm as an integrated development environment (IDE). The developed image processing algorithm was evaluated in both static and field conditions in two plots of chili plants. The developed system was able to capture a valid image of the chili plant under field conditions and accurately estimate the height and width of the plant with a RMSE in the ranges of 0.30-0.60 cm. The height and width measured by the proposed image processing algorithm were strongly correlated (R2 = 0.80–0.95) with manually measured values. Furthermore, the image processing approach has much more advanced features to measure the more complex geometric traits of plants.