We present RootPainter, a GUI-based software tool for the rapid training of deep neural networks for use in biological image analysis. RootPainter facilitates both fully-automatic and semiautomatic image segmentation. We investigate the effectiveness of RootPainter using three plant image datasets, evaluating its potential for root length extraction from chicory roots in soil, biopore counting and root nodule counting from scanned roots. We also use RootPainter to compare dense annotations to corrective ones which are added during the training based on the weaknesses of the current model. Deep Learning | GUI | Segmentation | Phenotyping | Biopore | Rhizotron | Root nodule | Interactive segmentation Correspondence: ags@di.ku.dk Fig. 1. RootPainter corrective annotation concept. (a) Roots in soil. (b) AI root predictions. (c) Human corrections. (d) AI learns from corrections. Smith et al. | bioRχiv | April 16, 2020 | 1-16
Taprooting crop species are capable of creating soil biopores (>2 mm in diameter) in the subsoil due to their large root size and deep-rooting habit. The aim of this study was to quantify root growth dynamics of wheat in the subsoil during its complete growth season as affected by crop sequence. Temporal observation on root length (km m −2 ) of wheat inside and outside of biopores at four growth stages (tillering, booting, anthesis, and milk) was conducted by using the profile wall method under the two crop sequence treatments involving two precrops, viz., chicory and tall fescue. Frequency of biopore presence measured on vertical profile walls depended on the choice of precrops in which chicory precrop resulted in higher frequency (2.3 %) compared with tall fescue (1.5 %). Root length of wheat measured inside biopores was significantly higher when grown after chicory (0.024 km m −2 ) in comparison to tall fescue (0.006 km m −2 ). On average, root length outside biopores after growing chicory was 45.9 % higher than tall fescue until the stage of anthesis. We conclude that at the site under study biopores as pathways for rapid root growth into deeper soil layers allow roots to re-enter and explore the subsoil. Thus, cereals cultivated in rotation with taprooted crops can draw benefit from enhanced uptake of water and nutrients from deeper soil layers during early growth stages. Model simulations with various abiotic and biotic factors will be helpful to reveal the direct evidence of biopore-root-shoot relationship in the future.
We present RootPainter, a GUI-based software tool for the rapid training of deep neural networks for use in biological image analysis. RootPainter facilitates both fully-automatic and semiautomatic image segmentation. We investigate the effectiveness of RootPainter using three plant image datasets, evaluating its potential for root length extraction from chicory roots in soil, biopore counting and root nodule counting from scanned roots. We also use RootPainter to compare dense annotations to corrective ones which are added during the training based on the weaknesses of the current model.
Depth and architecture of root systems play a prominent role in crop productivity under conditions of low water and nutrient availability. The subsoil contains high amounts of nutrients that may potentially serve for nutrient uptake by crops including finite resources such as phosphorus that have to be used in moderation to delay their exhaustion. Biopores are tubular shaped continuous soil pores formed by plant roots and earthworms. Taproot systems especially those of perennial legumes can make soil nutrients plant available from the solid phase and increase the density of vertical biopores in the subsoil thus making subsoil layers more accessible for succeeding crops. Density of larger sized biopores is further enhanced by increased abundance and activity of anecic earthworms resulting from soil rest and amount of provided feed. Nutrient rich drilospheres can provide a favorable environment for roots and nutrient uptake of subsequent crops. Future efficient nutrient management and crop rotation design in organic agriculture should entail these strategies of soil fertility building and biopore services in subsoil layers site specifically. Elements of these concepts are suggested to be used also in mainstream agriculture headlands, e.g. as 'Ecological Focus Areas', in order to improve soil structure as well as to establish a web of biodiversity while avoiding constraints for agricultural production.
The scale of root quantification in research is often limited by the time required for sampling, measurement and processing samples. Recent developments in Convolutional Neural Networks (CNN) have made faster and more accurate plant image analysis possible which may significantly reduce the time required for root measurement, but challenges remain in making these methods accessible to researchers without an in-depth knowledge of Machine Learning. We analyzed root images acquired from three destructive root samplings using the RootPainter CNN-software that features an interface for corrective annotation for easier use. Root scans with and without non-root debris were used to test if training a model, i.e., learning from labeled examples, can effectively exclude the debris by comparing the end-results with measurements from clean images. Root images acquired from soil profile walls and the cross-section of soil cores were also used for training and the derived measurements were compared with manual measurements. After 200 minutes of training on each dataset, significant relationships between manual measurements and RootPainter-derived data were noted for monolith (R2=0.99), profile wall (R2=0.76) and core-break (R2=0.57). The rooting density derived from images with debris was not significantly different from that derived from clean images after processing with RootPainter. Rooting density was also successfully calculated from both profile wall and soil core images, and in each case the gradient of root density with depth was not significantly different from manual counts. Our results demonstrate that the proposed approach using CNN can lead to substantial reductions in root sample processing workloads, increasing the potential scale of future root investigations.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.