Computer vision is increasingly used in farmers'fields and agricultural experiments to quantify important traits.Imaging setups with a sub-millimeter ground sampling distance enable the detection and tracking of p...Computer vision is increasingly used in farmers'fields and agricultural experiments to quantify important traits.Imaging setups with a sub-millimeter ground sampling distance enable the detection and tracking of plant features,including size,shape,and colour.Although today's AI-driven foundation models segment almost any object in an image,they still fail for complex plant canopies.To improve model performance,the global wheat dataset consortium assembled a diverse set of images from experiments around the globe.After the head detection dataset(GWHD),the new dataset targets a full semantic segmentation(GWFSS)of organs(leaves,stems and spikes)covering all developmental stages.Images were collected by 11 institutions using a wide range of imaging setups.Two datasets are provided:ⅰ)a set of 1096 diverse images in which all organs were labelled at the pixel level,and(ⅱ)a dataset of 52,078 images without annotations available for additional training.The labelled set was used to train segmentation models based on DeepLabV3Plus and Segformer.Our Segformer model performed slightly better than DeepLabV3Plus with a mIOU for leaves and spikes of ca.90%.However,the precision for stems with 54%was rather lower.The major advantages over published models are:ⅰ)the exclusion of weeds from the wheat canopy,ⅱ)the detection of all wheat features including necrotic and se-nescent tissues and its separation from crop residues.This facilitates further development in classifying healthy vs.unhealthy tissue to address the increasing need for accurate quantification of senescence and diseases in wheat canopies.展开更多
Phenomics technologies allow quantitative assessment of phenotypes across a larger number of plant genotypes compared to traditional phenotyping approaches.The utilization of such technologies has enabled the generati...Phenomics technologies allow quantitative assessment of phenotypes across a larger number of plant genotypes compared to traditional phenotyping approaches.The utilization of such technologies has enabled the generation of multidimensional plant traits creating big datasets.However,to harness the power of phenomics technologies,more sophisticated data analysis methods are required.In this study,Aphanomyces root rot(ARR)resistance in 547 lentil accessions and lines was evaluated using Red-Green-Blue(RGB)images of roots.We created a dataset of 6,460 root images that were annotated by a plant breeder based on the disease severity.Two approaches,generalized linear model with elastic net regularization(EN)and convolutional neural network(CNN),were developed to classify disease resistance categories into three classes:resistant,partially resistant,and susceptible.The results indicated that the selected image features using EN models were able to classify three disease categories with an accuracy of up to 0:91±0:004(0:96±0:005 resistant,0:82±0:009 partially resistant,and 0:92±0:007 susceptible)compared to CNN with an accuracy of about 0:84±0:009(0:96±0:008 resistant,0:68±0:026 partially resistant,and 0:83±0:015 susceptible).The resistant class was accurately detected using both classification methods.However,partially resistant class was challenging to detect as the features(data)of the partially resistant class often overlapped with those of resistant and susceptible classes.Collectively,the findings provided insights on the use of phenomics techniques and machine learning approaches to provide quantitative measures of ARR resistance in lentil.展开更多
Biogenic volatile organic compounds(VOCs)emitted by plants can reveal information about plant adaptation,defense processes,and biological pathways.Thus,such VOC data may be utilized to capture phenotypic plant respons...Biogenic volatile organic compounds(VOCs)emitted by plants can reveal information about plant adaptation,defense processes,and biological pathways.Thus,such VOC data may be utilized to capture phenotypic plant responses to the environment.In this study,the main objective was to evaluate the potential of biogenic compounds,including VOCs,to phenotype two pea cultivars,Ariel(susceptible)and Hampton(high levels of partial resistance)for resistance to Aphanomyces root rot disease.Plants were monitored non-destructively for VOC emission at three-time points(15,20,and 30 days after inoculation,DAI)using dynamic headspace sampling with gas chromatography-flame ionization detec-tion(GC-FID)system,as well as destructively at the end of the experiments,using solvent extraction and pyrolysis of both shoot and root tissues.A non-inoculated control(mock-inoculated with distilled water)was utilized to compare the plant responses within a cul-tivar.The common chemical peaks between control and inoculated samples of both culti-vars(RT_(cm))were analyzed after normalizing the relative peak intensity of inoculated samples with those of control samples,prior to a comparison between cultivars.In addi-tion,unique chemical peaks(RT_(uq))present in inoculated samples,but not in control sam-ples were also identified and their relative peak intensities were compared.Among the released green leaf volatiles(RT_(cm)),the normalized relative peak intensity of hexanal emis-sion,at 20 DAI,was higher in Ariel than that of Hampton.In addition,several putative chemical peaks(both RT_(cm) and RT_(uq)),previously known as indicators for disease response,exhibited some differences in their emission rates between pea cultivars in at least one of the time points.The destructive sampling revealed that shoot samples produced more putative unique biomarkers(RT_(uq))than the root samples.Based on the differences in puta-tive chemical peaks between cultivars,this initial study supports the concept of utilization of biogenic biomarker-based phenotyping in distinguishing levels of resistance in the eval-uated pea cultivars.More research is needed to further this approach for phenotyping other plant cultivars.Upon validation,the VOC profile integrated with high-throughput VOC sensing techniques can serve as a novel mechanism for phenotyping disease responses in crops.展开更多
基金Global wheat was directly supported by Analytics for the Australian Grains Industry(AAGI).
文摘Computer vision is increasingly used in farmers'fields and agricultural experiments to quantify important traits.Imaging setups with a sub-millimeter ground sampling distance enable the detection and tracking of plant features,including size,shape,and colour.Although today's AI-driven foundation models segment almost any object in an image,they still fail for complex plant canopies.To improve model performance,the global wheat dataset consortium assembled a diverse set of images from experiments around the globe.After the head detection dataset(GWHD),the new dataset targets a full semantic segmentation(GWFSS)of organs(leaves,stems and spikes)covering all developmental stages.Images were collected by 11 institutions using a wide range of imaging setups.Two datasets are provided:ⅰ)a set of 1096 diverse images in which all organs were labelled at the pixel level,and(ⅱ)a dataset of 52,078 images without annotations available for additional training.The labelled set was used to train segmentation models based on DeepLabV3Plus and Segformer.Our Segformer model performed slightly better than DeepLabV3Plus with a mIOU for leaves and spikes of ca.90%.However,the precision for stems with 54%was rather lower.The major advantages over published models are:ⅰ)the exclusion of weeds from the wheat canopy,ⅱ)the detection of all wheat features including necrotic and se-nescent tissues and its separation from crop residues.This facilitates further development in classifying healthy vs.unhealthy tissue to address the increasing need for accurate quantification of senescence and diseases in wheat canopies.
基金This activity was funded in part by US Department of Agriculture(USDA)-National Institute for Food and Agriculture(NIFA)Agriculture and Food Research Initiative Competitive Project WNP06825(accession number 1011741)Hatch Project WNP00011(accession number 1014919)the Washington State Department of Agriculture,Specialty Crop Block Grant program(project K1983).
文摘Phenomics technologies allow quantitative assessment of phenotypes across a larger number of plant genotypes compared to traditional phenotyping approaches.The utilization of such technologies has enabled the generation of multidimensional plant traits creating big datasets.However,to harness the power of phenomics technologies,more sophisticated data analysis methods are required.In this study,Aphanomyces root rot(ARR)resistance in 547 lentil accessions and lines was evaluated using Red-Green-Blue(RGB)images of roots.We created a dataset of 6,460 root images that were annotated by a plant breeder based on the disease severity.Two approaches,generalized linear model with elastic net regularization(EN)and convolutional neural network(CNN),were developed to classify disease resistance categories into three classes:resistant,partially resistant,and susceptible.The results indicated that the selected image features using EN models were able to classify three disease categories with an accuracy of up to 0:91±0:004(0:96±0:005 resistant,0:82±0:009 partially resistant,and 0:92±0:007 susceptible)compared to CNN with an accuracy of about 0:84±0:009(0:96±0:008 resistant,0:68±0:026 partially resistant,and 0:83±0:015 susceptible).The resistant class was accurately detected using both classification methods.However,partially resistant class was challenging to detect as the features(data)of the partially resistant class often overlapped with those of resistant and susceptible classes.Collectively,the findings provided insights on the use of phenomics techniques and machine learning approaches to provide quantitative measures of ARR resistance in lentil.
基金This activity was funded in part by the U.S.Department of Agriculture-National Institute for Food and Agriculture(USDA-NIFA)Agriculture and Food Research Initiative(AFRI)Competitive Project WNP06825(ac-cession number 1011741)Hatch Project WNP00011(accession number 1014919)CAHNRS Emerging Research Issues project.
文摘Biogenic volatile organic compounds(VOCs)emitted by plants can reveal information about plant adaptation,defense processes,and biological pathways.Thus,such VOC data may be utilized to capture phenotypic plant responses to the environment.In this study,the main objective was to evaluate the potential of biogenic compounds,including VOCs,to phenotype two pea cultivars,Ariel(susceptible)and Hampton(high levels of partial resistance)for resistance to Aphanomyces root rot disease.Plants were monitored non-destructively for VOC emission at three-time points(15,20,and 30 days after inoculation,DAI)using dynamic headspace sampling with gas chromatography-flame ionization detec-tion(GC-FID)system,as well as destructively at the end of the experiments,using solvent extraction and pyrolysis of both shoot and root tissues.A non-inoculated control(mock-inoculated with distilled water)was utilized to compare the plant responses within a cul-tivar.The common chemical peaks between control and inoculated samples of both culti-vars(RT_(cm))were analyzed after normalizing the relative peak intensity of inoculated samples with those of control samples,prior to a comparison between cultivars.In addi-tion,unique chemical peaks(RT_(uq))present in inoculated samples,but not in control sam-ples were also identified and their relative peak intensities were compared.Among the released green leaf volatiles(RT_(cm)),the normalized relative peak intensity of hexanal emis-sion,at 20 DAI,was higher in Ariel than that of Hampton.In addition,several putative chemical peaks(both RT_(cm) and RT_(uq)),previously known as indicators for disease response,exhibited some differences in their emission rates between pea cultivars in at least one of the time points.The destructive sampling revealed that shoot samples produced more putative unique biomarkers(RT_(uq))than the root samples.Based on the differences in puta-tive chemical peaks between cultivars,this initial study supports the concept of utilization of biogenic biomarker-based phenotyping in distinguishing levels of resistance in the eval-uated pea cultivars.More research is needed to further this approach for phenotyping other plant cultivars.Upon validation,the VOC profile integrated with high-throughput VOC sensing techniques can serve as a novel mechanism for phenotyping disease responses in crops.
