Transfer Learning from Synthetic Data Applied to Soil–Root Segmentation in X-Ray Tomography Images

Douarre, Clément; Schielein, Richard; Frindel, Carole; Gerth, Stefan; Rousseau, D.

doi:10.3390/jimaging4050065

Cited by 60 publications

(36 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…CNNs have now established their dominance on almost all recognition and detection tasks [42][43][44][45]. They have also been used to segment roots from soil in X-ray tomography [46] and to identify the tips of wheat roots grown in germination paper growth pouches [41]. CNNs have an ability to transfer well from one task to another, requiring less training data for new tasks [47].…”

Section: Open Accessmentioning

confidence: 99%

Segmentation of roots in soil with U-Net

et al. 2020

View full text Add to dashboard Cite

Background Plant root research can provide a way to attain stress-tolerant crops that produce greater yield in a diverse array of conditions. Phenotyping roots in soil is often challenging due to the roots being difficult to access and the use of time consuming manual methods. Rhizotrons allow visual inspection of root growth through transparent surfaces. Agronomists currently manually label photographs of roots obtained from rhizotrons using a line-intersect method to obtain root length density and rooting depth measurements which are essential for their experiments. We investigate the effectiveness of an automated image segmentation method based on the U-Net Convolutional Neural Network (CNN) architecture to enable such measurements. We design a data-set of 50 annotated chicory (Cichorium intybus L.) root images which we use to train, validate and test the system and compare against a baseline built using the Frangi vesselness filter. We obtain metrics using manual annotations and line-intersect counts. Results Our results on the held out data show our proposed automated segmentation system to be a viable solution for detecting and quantifying roots. We evaluate our system using 867 images for which we have obtained line-intersect counts, attaining a Spearman rank correlation of 0.9748 and an $$r^2$$r2 of 0.9217. We also achieve an $$F_1$$F1 of 0.7 when comparing the automated segmentation to the manual annotations, with our automated segmentation system producing segmentations with higher quality than the manual annotations for large portions of the image. Conclusion We have demonstrated the feasibility of a U-Net based CNN system for segmenting images of roots in soil and for replacing the manual line-intersect method. The success of our approach is also a demonstration of the feasibility of deep learning in practice for small research groups needing to create their own custom labelled dataset from scratch.

show abstract

Section: Open Accessmentioning

confidence: 99%

Segmentation of roots in soil with U-Net

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Main considerations include sufficient amount of balanced data, annotation and normalization of data, and outlier rejection [5]. Synthetic data modelling, graphical rendering, and transfer learning in context of using pre-trained deep networks (or at least their first layers) for various tasks is also mentioned in other papers dealing with plant genotyping and phenotyping [6].…”

Section: Introductionmentioning

confidence: 99%

Data Augmentation for Leaf Segmentation and Counting Tasks in Rosette Plants

Kuznichov

Zvirin

Honen

et al. 2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)

View full text Add to dashboard Cite

Deep learning techniques involving image processing and data analysis are constantly evolving. Many domains adapt these techniques for object segmentation, instantiation and classification. Recently, agricultural industries adopted those techniques in order to bring automation to farmers around the globe. One analysis procedure required for automatic visual inspection in this domain is leaf count and segmentation. Collecting labeled data from field crops and greenhouses is a complicated task due to the large variety of crops, growth seasons, climate changes, phenotype diversity, and more, especially when specific learning tasks require a large amount of labeled data for training. Data augmentation for training deep neural networks is well established, examples include data synthesis, using generative semi-synthetic models, and applying various kinds of transformations. In this paper we propose a method that preserves the geometric structure of the data objects, thus keeping the physical appearance of the data-set as close as possible to imaged plants in real agricultural scenes. The proposed method provides state of the art results when applied to the standard benchmark in the field, namely, the ongoing Leaf Segmentation Challenge hosted by Computer Vision Problems in Plant Phenotyping.

show abstract

“…Obtaining labeled data from real RSA is difficult, error-prone, and time-consuming. Inspired by [19,8], we use publicly available synthetic root data that is artificially generated instead to train our model [18]. This synthetic dataset contains dicot and monocot roots of large resolution (see Figure 2(a)).…”

Section: Dataset and Augmentationmentioning

confidence: 99%

Adversarial Large-Scale Root Gap Inpainting

Chen

Giuffrida

Doerner

et al. 2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)

View full text Add to dashboard Cite

Root imaging of a growing plant in a non-invasive, affordable, and effective way remains challenging. One approach is to image roots by growing them in a rhizobox, a soil-filled transparent container, imaging them with digital cameras, and segmenting root from soil background. However, due to soil occlusion and the fact that digital imaging is a 2D projection of a 3D object, gaps are present on the segmentation masks, which may hinder the extraction of finely grained root system architecture (RSA) traits. Herein, we develop an image inpainting technique to recover gaps from disconnected root segments. We train a patch-based deep fully convolutional network using a supervised loss but also use adversarial mechanisms at patch and whole root level. We use Policy Gradient method, to endow the model with large-scale whole root view during training. We train our model using synthetic root data. In our experiments, we show that using adversarial mechanisms at local and whole-root level we obtain a 72% improvement in performance on recovering gaps of real chickpea data when using only patch-level supervision.

show abstract

Transfer Learning from Synthetic Data Applied to Soil–Root Segmentation in X-Ray Tomography Images

Cited by 60 publications

References 31 publications

Segmentation of roots in soil with U-Net

Segmentation of roots in soil with U-Net

Data Augmentation for Leaf Segmentation and Counting Tasks in Rosette Plants

Adversarial Large-Scale Root Gap Inpainting

Contact Info

Product

Resources

About