Distortion Correction in 3D-Modeling of Root Systems for Plant Phenotyping

Tushar Kanta Das Nakini, Guilherme N. DeSouza

Abstract: Root Phenotyping is an important tool in predicting the life and growth of plants. Many systems have been developed to automate the process of extracting root traits using 3D imaging system, however, not many of those systems corrected for the distortions that frequently appear during this process. In this paper we present a new method to compensate for light refractions that occur due to hydroponic substrates – gel-based platforms for growing plants. As our results demonstrate, our method provides an accurate 3D point cloud containing the coordinates of the surface of the root system with error smaller than 0.16 mm in average and standard deviation of less than 0.13 mm.