Development of methods to quantify root distribution, length and biomass using a machine learning approach based on X-ray tomography of intact field soil cores

Growers want to improve their soil management practices to improve the health and productivity of their soils. A key factor in determining the impact of soil management practices is knowing the effect of the management practice on the crop's root systems because it is through the roots that crops access water and nutrients that help them grow. However, soil-root interactions happen underground making it difficult to determine the impact of soil management practices on root growth and distribution. Recent research at the Australian Synchrotron has shown that root distribution and growth in large intact soil cores can be successfully analysed using x-ray computed tomography. While effective, this is a time-consuming and expensive process that requires high performance computing, which is not readily available. This investment will explore machine learning as a possible way to reduce the time to analyse samples through the use of more widely accessible computers. The investment will also explore other cheaper and more accessible approaches to analyse roots, such as conventional (benchtop) x-ray computed tomography imaging facilities and this investment also includes a top-up scholarship for a PhD candidate to work with the University of South Australia on the project.