The Terrestrial Laser Scanner for Cassava (Manihot esculenta) Phenotyping & the Development of a Low-Cost 3D Phenotyping Platform

Abstract

There will be a large increase in the world’s population over the coming decades, which will require a significant increase in global food production. However, a large gap exists between the cultivars we plant today and those required to feed the world of tomorrow. We currently lack the ability to rapidly phenotype for those improved cultivars, causing a bottleneck in the breeding process. One tool which might speed up the phenotyping process is LIDAR, or light detection and ranging. This technology has proven useful in the phenotyping of many characteristics across a number of crops; however, it has not been demonstrated on cassava. Cassava is of global importance as both a source of starch as well as a foodstuff for cattle and, to a lesser extent, as a leafy vegetable. We demonstrate the applicability of using terrestrial LIDAR for the determination of cassava height, as well as total aboveground biomass and total leaf biomass, all within a field setting. In addition, we showcase the development of a novel, low-cost 3D phenotyping platform designed around Microsoft Kinect V2 sensors. The platform was designed to combat the high acquisition costs generally associated with, and limiting of, the study of 3D high-throughput phenotyping.