Non-destructive 3d geometric modeling of maize root-stubble in-situ via x-ray computed tomography

Abstract

No-tillage seeding has become an important approach to improve crop productivity, which needs colters of high performance to cut the root-stubble-soil composite. However, the difficulty of maize root-stubbles three-dimensional (3D) modeling hinders finite element (FE) simulation to improve development efficiency of such colters because of maize root system complexity and opaque nature of the soil. Fortunately, the non-destructive 3D geometric model of the maize root-stubble in-situ can be established via X-ray computed tomography (CT) following by a systematic procedure. The whole procedure includes CT scanning of the maize root-stubble-soil composite sample, image reconstruction via filtered back-projection (FBP) with the Hanning filter, segmentation of root-stubble via a variational level set method, and post-processing via morphological operations. The 3D reconstruction model of the maize root-stubble in-situ presents a complete, complex and in-situ geometrical morphology, which cannot be realized via other methods, including the destructive modelling after washing via CT. This study is the first to build a 3D geometric model of a maize root-stubble in-situ via CT, which opens up new possibilities for simulation of root-stubble-soil cutting using FEM, and much other research related to plant root-stubbles.