Dataset Information

An Exploration of Deep-Learning Based Phenotypic Analysis to Detect Spike Regions in Field Conditions for UK Bread Wheat.

ABSTRACT: Wheat is one of the major crops in the world, with a global demand expected to reach 850 million tons by 2050 that is clearly outpacing current supply. The continual pressure to sustain wheat yield due to the world's growing population under fluctuating climate conditions requires breeders to increase yield and yield stability across environments. We are working to integrate deep learning into field-based phenotypic analysis to assist breeders in this endeavour. We have utilised wheat images collected by distributed CropQuant phenotyping workstations deployed for multiyear field experiments of UK bread wheat varieties. Based on these image series, we have developed a deep-learning based analysis pipeline to segment spike regions from complicated backgrounds. As a first step towards robust measurement of key yield traits in the field, we present a promising approach that employ Fully Convolutional Network (FCN) to perform semantic segmentation of images to segment wheat spike regions. We also demonstrate the benefits of transfer learning through the use of parameters obtained from other image datasets. We found that the FCN architecture had achieved a Mean classification Accuracy (MA) >82% on validation data and >76% on test data and Mean Intersection over Union value (MIoU) >73% on validation data and and >64% on test datasets. Through this phenomics research, we trust our attempt is likely to form a sound foundation for extracting key yield-related traits such as spikes per unit area and spikelet number per spike, which can be used to assist yield-focused wheat breeding objectives in near future.

SUBMITTER: Alkhudaydi T

PROVIDER: S-EPMC7706304 | biostudies-literature | 2019

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

An Exploration of Deep-Learning Based Phenotypic Analysis to Detect Spike Regions in Field Conditions for UK Bread Wheat.

Alkhudaydi Tahani T Reynolds Daniel D Griffiths Simon S Zhou Ji J de la Iglesia Beatriz B

Plant phenomics (Washington, D.C.) 20190731

Wheat is one of the major crops in the world, with a global demand expected to reach 850 million tons by 2050 that is clearly outpacing current supply. The continual pressure to sustain wheat yield due to the world's growing population under fluctuating climate conditions requires breeders to increase yield and yield stability across environments. We are working to integrate deep learning into field-based phenotypic analysis to assist breeders in this endeavour. We have utilised wheat images col ...[more]

PMID: 33313535

Similar Datasets

Project description:Increasing grain quality and nutritional value along with yield in bread wheat is one of the leading breeding goals. Selection of genotypes with desired traits using traditional breeding selection methods is very time-consuming and often not possible due to the interaction of environmental factors. By identifying DNA markers that can be used to identify genotypes with desired alleles, high-quality and bio-fortified bread wheat production can be achieved in a short time and cost-effectively. In the present study, 134 doubled haploid (DH) wheat lines and their four parents were phenotypically evaluated for yield components (spike characteristics), quality parameters, and grain Fe and Zn concentrations in two successive growing seasons. At the same time, ten genic simple sequence repeats (SSR) markers linked to genes related to the traits examined were validated and subsequently used for molecular characterization of trait-specific candidate genotypes. Significant genotypic variations were determined for all studied traits and many genotypes with desired phenotypic values were detected. The evaluation performed with 10 SSR markers revealed significant polymorphism between genotypes. The polymorphic information content (PIC) values of 10 markers ranged from 0.00 to 0.87. Six out of 10 SSRs could be more effective in representing the genotypic differentiation of the DH population as they demonstrated the highest genetic diversity. Both Unweighted Pair Group Method with Arithmetic Mean (UPGMA) clustering and STRUCTURE analyses divided 138 wheat genotypes into five (K = 5) main groups. These analyzes were indicative of genetic variation due to hybridization and segregation in the DH population and the differentiation of the genotypes from their parents. Single marker regression analysis showed that both Xbarc61 and Xbarc146 had significant relationships with grain Fe and Zn concentrations, while Xbarc61 related to spike characteristics and Xbarc146 related to quality traits, separately. Other than these, Xgwm282 was associated with spike harvest index, SDS sedimentation value and Fe grain concentration, while Gwm445 was associated with spikelet number, grain number per spike and grain Fe concentration. These markers were validated for the studied DH population during the present study and they could be effectively used for marker-assisted selection to improve grain yield, quality, and bio-fortification capacity of bread wheat.

Project description:BackgroundUnderstanding root system morphology in bread wheat is critical for identifying root traits to breed cultivars with improved resource uptake and better adaptation to adverse environments. Variability in root morphological traits at early vegetative stages was examined among 184 bread wheat genotypes originating from 37 countries grown in a semi-hydroponic phenotyping system.ResultsAt the onset of tillering (Z2.1, 35?days after transplanting), plants had up to 42?cm in shoot height and 158?cm long in root depth. Phenotypic variation existed for both shoot and root traits, with a maximal 4.3-fold difference in total root length and 5-fold difference in root dry mass among the 184 genotypes. Of the 41 measured traits, 24 root traits and four shoot traits had larger coefficients of variation (CV???0.25). Strong positive correlations were identified for some key root traits (i.e., root mass, root length, and these parameters at different depths) and shoot traits (i.e., shoot mass and tiller number) (P???0.05). The selected 25 global traits (at whole-plant level) contributed to one of the five principal components (eigenvalues>?1) capturing 83.0% of the total variability across genotypes. Agglomerative hierarchical clustering analysis separated the 184 genotypes into four (at a rescaled distance of 15) or seven (at a rescaled distance of 10) major groups based on the same set of root traits. Strong relationships between performance traits (dry mass) with several functional traits such as specific root length, root length intensity and root tissue density suggest their linkage to plant growth and fitness strategies.ConclusionsLarge phenotypic variability in root system morphology in wheat genotypes was observed at the tillering stage using established semi-hydroponic phenotyping techniques. Phenotypic differences in and trait correlations among some interesting root traits may be considered for breeding wheat cultivars with efficient water acquisition and better adaptation to abiotic stress.

Dataset Information

An Exploration of Deep-Learning Based Phenotypic Analysis to Detect Spike Regions in Field Conditions for UK Bread Wheat.

Publications

An Exploration of Deep-Learning Based Phenotypic Analysis to Detect Spike Regions in Field Conditions for UK Bread Wheat.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets