Modeling Cotton Growth and Nitrogen Status Using Image Analysis

Abstract

Core Ideas The study developed an image analysis approach for extracting the color features of a cotton canopy with a digital camera. The study verified the models using data collected from three fields of high‐yielding cotton. Image analysis techniques have been widely used to monitor crop growth and diagnose crop N status. The main objective of this study was to develop an image analysis method for monitoring cotton (Gossypium hirsutum L.) growth and N status with images taken by a digital camera. The experiment was consisted of two cotton cultivars grown under five different N application rates. Ten color indices were extracted from images of the cotton canopy taken between emergence and full bloom. Results showed that nine out of ten color indices were significantly correlated with N content, leaf area index (LAI), and dry matter accumulation (DMA). Among the color indices, the green minus red index (GMRI), excess green index (EGI), green to red index (GDRI), and canopy cover (CC) had the highest correlation with the three crop properties. Models were developed to describe the relationships between the color indices (GMRI, EGI, GDRI, and CC) and the crop properties (N content, LAI, and DMA). Relationships between the color indices and the three crop properties were exponential. The models were tested in three large and representative high‐yielding cotton fields. Results showed that GMRI, EGI, GDRI, and CC are valid indicators for monitoring cotton growth and N status. Digital cameras have a good potential for monitoring cotton growth and N status.