Abstract
The Rhizotrons method is an important means of detecting dynamic growth and development phenotypes of plant roots. However, the segmentation of root images is a critical obstacle restricting further development of this method. At present, researchers mostly use direct manual drawings or software-assisted manual drawings to segment root systems for analysis. Root systems can be segmented from root images obtained by the Rhizotrons method, and then, root system lengths and diameters can be obtained with software. This type of image segmentation method is extremely inefficient and very prone to human error. Here, we investigate the effectiveness of an automated image segmentation method based on the DeepLabv3+ convolutional neural network (CNN) architecture to streamline such measurements. We have improved the upsampling portion of the DeepLabv3+ network and validated it using in situ images of cotton roots obtained with a micro root window root system monitoring system. Segmentation performance of the proposed method utilizing WinRHIZO Tron MF analysis was assessed using these images. After 80 epochs of training, the final verification set F1-score, recall, and precision were 0.9773, 0.9847, and 0.9702, respectively. The Spearman rank correlation between the manually obtained Rhizotrons manual segmentation root length and automated root length was 0.9667 (p < 10–8), with r2 = 0.9449. Based on the comparison of our segmentation results with those of traditional manual and U-net segmentation methods, this novel method can more accurately segment root systems in complex soil environments. Thus, using the improved DeepLabv3+ to segment root systems based on micro-root images is an effective method for accurately and quickly segmenting root systems in a homogeneous soil environment and has clear advantages over traditional manual segmentation.
Highlights
The growth environment of plant roots within the soil is extremely complex
DeepLabv3+ achieved precision and recall values of 0.9702 and 0.9847, respectively, with the validation set, which means that the number of pixels in the model that mistake the soil background for the root is greater than the number of pixels that mistake the root for the soil background
The high-resolution in situ root images collected by the Rhizotrons method generally segments roots and obtains root morphological indicators using WinRHIZO Tron MF software, which is a traditional manual segmentation method (Munoz-Romero et al, 2010)
Summary
The growth environment of plant roots within the soil is extremely complex. soil is a non-transparent medium, so it is difficult to quickly and accurately obtain phenotypic information, which is a critical obstacle to research on root development. These destructive sampling methods do not enable phenotypic observations of dynamic root systems in situ To address this limitation, non-destructive observation methods such as X-ray computed tomography (CT) (Mairhofer et al, 2012, 2013; Mooney et al, 2012), nuclear magnetic resonance (NMR) imaging technology (Pflugfelder et al, 2017), laser scanning (Fang et al, 2009), and 3D imaging (Iyer-Pascuzzi et al, 2010; Clark et al, 2011; Topp et al, 2013) have been applied. The quality of in situ image segmentation underlies the quality of the root phenotype results
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
