Spectral remote sensing allows estimation of important biochemical and biophysical crop traits such as chlorophyll and nitrogen content, leaf area, and biomass. At the same time it also has limitations, for example, due to challenges that arise from nonconstant illumination conditions or saturation effects at high leaf area index. Motivated by findings that showed that plant height is also a good predictor for crop biomass, Three-dimensional (3D) data has gained attention for crop trait extraction in recent years. This chapter gives an overview of the approach for extracting crop traits from 3D data as well as from combined 3D and spectral sensing.
Terrestrial laser scanning or images from cameras mounted on unmanned aerial vehicles in combination with the “structure from motion” technique are two increasingly popular approaches to derive 3D information of a crop stand. Three-dimensional data can be used to estimate 3D crop traits such as crop height or other structural parameters of the canopy. Multitemporal campaigns allow one to follow the crop growth. Additionally, spectral 2D imagers allow capture of 3D and spectral information at the same time.
Several approaches exist to use 3D and spectral data for crop trait extraction. In a segmentation approach, either spectral (or color) or 3D information is used for presegmentation or classification of the data. A complementary approach uses spectral and 3D data as complementary data to estimate different traits from both type of data, and a combination approach combines 3D and spectral data to estimate one trait. These approaches have already provided promising results and we think that 3D and spectral data will increasingly be used together to extract crop traits in support for phenotyping and precision agriculture applications.