SeedGerm: a cost‐effective phenotyping platform for automated seed imaging and machine‐learning based phenotypic analysis of crop seed germination

Colmer, Joshua; O’Neill, Carmel M.; Wells, Rachel; Bostrom, Aaron; Reynolds, Daniel B.; Websdale, Danny; Shiralagi, Gagan; Wei, Lu; Lou, Qiaojun; Cornu, Thomas Le; Ball, Joshua; Renema, Jim; Andaluz, Gema Flores; Benjamins, René; Penfield, Steven; Zhou, Ji

doi:10.1111/nph.16736

Cited by 87 publications

(48 citation statements)

References 68 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The high-throughput capabilities of this system are well suited to this type of research. A similar approach has been reported in our recent work, SeedGerm (Colmer et al, 2020), which was applied to detect genetic differences in Brassica napus based on a range of germination traits. Although more work is needed, greater automation of phenotypic analysis and improvements in accuracy are likely to accelerate genetic analysis of crop performance under varied treatments or environments.…”

Section: Discussionmentioning

confidence: 98%

Large-scale field phenotyping using backpack LiDAR and GUI-based CropQuant-3D to measure structural responses to different nitrogen treatments in wheat

Zhu¹,

Sun²,

G³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Plant phenomics is widely recognised as a key area to bridge the gap between traits of agricultural importance and genomic information. A wide range of field-based phenotyping solutions have been developed. Nevertheless, disadvantages of these current systems have been identified concerning mobility, affordability, accuracy, scalability, and the ability to analyse big data collected. Here, we present a novel solution that combines a commercial backpack LiDAR device and graphical user interface (GUI) based software called CropQuant-3D, which has been applied to analyse 3D morphological traits in wheat. To our knowledge, this is the first use of backpack LiDAR in field-based plant research, acquiring millions of 3D points to represent spatial features of crops. A key part of the innovation is the GUI-based software that can extract plot-based traits from large and complex point clouds with limited computing time. We describe how we developed and used the combined system to quantify canopy structural changes, impossible to measure previously. Also, we demonstrate the biological relevance of our work through a case study that examined wheat varieties to three different levels of nitrogen fertilisation in field experiments. The results indicate that the solution can differentiate significant genotype and treatment effects on key traits, with strong correlations with manual measurements. Hence, we believe that the solution presented here could consistently and speedily quantify traits at a larger scale, indicating the system could be used as a reliable research tool in large-scale and multi-location field phenotyping to contribute to the resolution of the phenotyping bottleneck.

show abstract

Section: Discussionmentioning

confidence: 98%

Large-scale field phenotyping using backpack LiDAR and GUI-based CropQuant-3D to measure structural responses to different nitrogen treatments in wheat

Zhu¹,

Sun²,

G³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…It produces curves based on seed-level germination timing and rates rather than a fitted curve that matches specialist scoring of radical emergences (Colmer et al, 2020).…”

Section: High Throughput Phenotyping and Software To Monitor Variatiomentioning

confidence: 99%

First the seed: Genomic advances in seed science for improved crop productivity and food security

Dwivedi¹,

Spillane

Lopez

et al. 2021

Crop Science

View full text Add to dashboard Cite

Seeds are valuable sources of carbohydrates, lipids, proteins, fibers, minerals, and vitamins. They provide energy and nutrition to germinating seedlings, food to humans, feed to livestock and feedstocks to industry. High-throughput analyses of gene expression in crops has identified many candidate genes associated with seed dormancy, longevity, germination, and vigor. In this review, we cover the latest research focusing on such key seed traits. Transcriptome analyses of time-courses of seed filling have identified sets of genes expressed at different stages of this process. The potential role of epigenetics (including seed-endosperm imprinted genes) in regulating seed development and chemistry is highlighted herein. We also discuss how advances in genomics and seed biology are facilitating the unravelling of associations between seed traits with genebank accessions and gene sequences, including how functional research can accelerate the discovery of allelic variants. Such knowledge of functional effects relating to gene variants is necessary for more efficient and cost-effective management of genetic resources or for redesigning crops with specific seed characteristics. For instance, genebank curators may assess seed viability by monitoring changes in gene expression of biomarker genes in dry seed samples to decide germplasm regeneration and assess genetic integrity of collections by monitoring changes in This article is protected by copyright. All rights reserved. diversity and allele frequencies between samples of same accession stored in genebanks. We highlight that resistance to preharvest sprouting can be enhanced through genomicsassisted breeding in otherwise nondormant rice and wheat cultivars, while pimt, another valuable marker for seed longevity, may be deployed to enhance seed vigor in crops.

show abstract

“…González et al (2020) developed MyROOT 2.0, which uses an automatic and efficient algorithm to detect the root regions of images; this improved the previous version MyROOT, which required manual intervention by the user to define the root area pattern (Betegón-Putze et al, 2019), and also improved the efficiency of batch root detection. Colmer et al (2020) proposed the SeedGerm system, which integrates automatic seed imaging and machine learning-based phenotype analysis, thus providing a wide range of applications for large-scale phenotype analysis and detection of plant seeds.…”

Section: Introductionmentioning

confidence: 99%

High-Throughput in situ Root Image Segmentation Based on the Improved DeepLabv3+ Method

et al. 2020

View full text Add to dashboard Cite

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 r 2 = 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.

show abstract

SeedGerm: a cost‐effective phenotyping platform for automated seed imaging and machine‐learning based phenotypic analysis of crop seed germination

Cited by 87 publications

References 68 publications

Large-scale field phenotyping using backpack LiDAR and GUI-based CropQuant-3D to measure structural responses to different nitrogen treatments in wheat

Large-scale field phenotyping using backpack LiDAR and GUI-based CropQuant-3D to measure structural responses to different nitrogen treatments in wheat

First the seed: Genomic advances in seed science for improved crop productivity and food security

High-Throughput in situ Root Image Segmentation Based on the Improved DeepLabv3+ Method

Contact Info

Product

Resources

About