As good as human experts in detecting plant roots in minirhizotron images but efficient and reproducible: the convolutional neural network “RootDetector”

Abstract

Plant roots influence many ecological and biogeochemical processes, such as carbon, water and nutrient cycling. Because of difficult accessibility, knowledge on plant root growth dynamics in field conditions, however, is fragmentary at best. Minirhizotrons, i.e. transparent tubes placed in the substrate into which specialized cameras or circular scanners are inserted, facilitate the capture of high-resolution images of root dynamics at the soil-tube interface with little to no disturbance after the initial installation. Their use, especially in field studies with multiple species and heterogeneous substrates, though, is limited by the amount of work that subsequent manual tracing of roots in the images requires. Furthermore, the reproducibility and objectivity of manual root detection is questionable. Here, we use a Convolutional Neural Network (CNN) for the automatic detection of roots in minirhizotron images and compare the performance of our RootDetector with human analysts with different levels of expertise. Our minirhizotron data come from various wetlands on organic soils, i.e. highly heterogeneous substrates consisting of dead plant material, often times mainly roots, in various degrees of decomposition. This may be seen as one of the most challenging soil types for root segmentation in minirhizotron images. RootDetector showed a high capability to correctly segment root pixels in minirhizotron images from field observations (F1 = 0.6044; r2 compared to a human expert = 0.99). Reproducibility among humans, however, depended strongly on expertise level, with novices showing drastic variation among individual analysts and annotating on average more than 13-times higher root length/cm2 per image compared to expert analysts. CNNs such as RootDetector provide a reliable and efficient method for the detection of roots and root length in minirhizotron images even from challenging field conditions. Analyses with RootDetector thus save resources, are reproducible and objective, and are as accurate as manual analyses performed by human experts.

The hidden roots of wetland methane emissions

“…X-ray computed tomography could be a useful tool for examining root system architecture and its potential influence on C substrate provision, gas transport and microbial distribution (Handakumbura et al, 2021;Hou et al, 2022;Yang et al, 2017). Minirhizotrons, together with improved artificial intelligence methods for imagebased root-detection, have been successfully applied in wetlands (Arnaud et al, 2021;Defrenne et al, 2021;D'Imperio et al, 2018;Iversen et al, 2012;Peters et al, 2023;Rodgers et al, 2004;Sciumbata et al, 2023).…”

Improved Transformer for Time Series Senescence Root Recognition

“…For instance, better UNet models were used to build root detection tools like ChronoRoot, RootPainter, RootDetector, etc. [ 32 – 34 ]. Pyramid pooling, on which PSPNet is built, aggregates the context data of many receptive fields and enhances the ability to gather global information [ 35 ].…”

A trainable object finder, selector and identifier for pollen, spores and other things: A step towards automated pollen recognition in lake sediments

“…In a recent collaboration within the DIG-IT! project, ecologists from the University of Greifswald and visual computing experts from the Fraunhofer Institute IGD in Rostock have explored the potential of automated image recognition for the study of wood anatomy, root growth, and bat monitoring (Gillert et al, 2023; Krivek et al, 2023; Peters et al, 2023; Resente et al, 2021). As a fourth focus, we investigated the automated recognition of subfossil pollen and spores in lake sediments.…”