Skip to main content

Table 1 Performance of different techniques/models (see Table 4 for details) on the homogeneous Zea mays (ATTRACT 1) validation set

From: Semantic segmentation of plant roots from RGB (mini-) rhizotron images—generalisation potential and false positives of established methods and advanced deep-learning models

Technique/model

SSIM

DSC

IoU

Dummy classifier

0.9154

0.0032

0.0032

Frangi Vesselness

0.4222

0.2634

0.1639

Adaptive thresholding

0.8076

0.5033

0.3664

SVM

0.8326

0.5804

0.4271

SegRoot

0.9203

0.4460

0.3122

UNetGNRes

0.9498

0.6708

0.5380

U-Net SE-ResNeXt-101 (32 × 4d)

0.9534

0.6968

0.5607

U-Net EfficientNet-b6

0.9551

0.7213

0.5860

  1. The best scores for the evaluation metrics average structural similarity index (SSIM), average Sørensen–Dice similarity coefficient (DSC), and average Jaccard index/Intersection over Union (IoU) are given in bold
Back to article page

Plant Methods

ISSN: 1746-4811