GPhenoVision: A ground mobile system with multi-modal imaging for field-based high throughput phenotyping of cotton

Abstract

Imaging sensors can extend phenotyping capability, but they require a system to handle high-volume data. The overall goal of this study was to develop and evaluate a field-based high throughput phenotyping system accommodating high-resolution imagers. The system consisted of a high-clearance tractor and sensing and electrical systems. The sensing system was based on a distributed structure, integrating environmental sensors, real-time kinematic GPS, and multiple imaging sensors including RGB-D, thermal, and hyperspectral cameras. Custom software was developed with a multilayered architecture for system control and data collection. The system was evaluated by scanning a cotton field with 23 genotypes for quantification of canopy growth and development. A data processing pipeline was developed to extract phenotypes at the canopy level, including height, width, projected leaf area, and volume from RGB-D data and temperature from thermal images. Growth rates of morphological traits were accordingly calculated. The traits had strong correlations (r = 0.54-0.74) with fiber yield and good broad sense heritability (H 2 = 0.27-0.72), suggesting the potential for conducting quantitative genetic analysis and contributing to yield prediction models. The developed system is a useful tool for a wide range of breeding/genetic, agronomic/physiological, and economic studies.