AAA・Efficientfruitselectionbynon-destructiveevaluationoffruitqualityinpost-harvestmanagement・Phenotypingoffruittraitsusingimagedataanddeeplearning・QuantitativeanalysisoforchardsusingimagedataforhighqualityfruitproductionResearch keywords:・Fruitand fruit tree・Nondestructiveevaluation・Image phenotyping・Deep learningResearch keywordsMillet species, SoybeanDrought, Salt, Waterlogging Grain yield, RootEcophisiology“Pomology”isabranchofhorticulturalsciencethatfocusesonfruitandfruittrees.Thisfieldinvolvesbasicresearchintophysiologyoffruitandfruittrees,aswellasappliedresearchnecessaryfortheefficientfruitproduction.Inparticular,ourresearchaimstodevelopnewapproachestoefficientfruitproduction.Wearegoingtodevelopnon-destructivephenotyping(AI:ArtificialIntelligence),basedonfruitandorchardimages.Wealsoaimtoquantifyandanalyzecultivationtechniquesthatenablethehighqualityfruitproductionimagedataanalysis.techniquesInrecentyears,globalwarminganddesertificationhavebecomemoreseriousduetohumanactivities,andextremeweatherhashadagreatimpactonagriculture.Inthehillyandmountainousareas,thereisamagnificentnaturethatfascinatespeople,anditisexpectedtobeactivatedbytakingadvantageofthecharacteristicsoftheareasuchastheproductionofspecialcrops,however,thegrainyieldislowerthanonflatlandandlarge-scalemanagementisdifficult.Thiscausedabandonedcultivatedlandisincreasing.Milletssuchasfoxtailmillet,commonmillet,andJapanesemilletareoneoftheoldestcrops,andhavebeenanimportantfoodsourcesinceancienttimes.Theyareresistanttosoildryness,excessivehumidity,saltdamage,etc.However,therearefewstudiesonmilletspecies.Wecollectmilletandsoybeanspeciesfromallovertheworldandinvestigateindetailtheadaptiveresponsetoenvironmentalstressandthecharacteristicsofcerealsthatareresistanttobirdandpests.“What is the mechanism of millet that grows"To establish sustainable agriculture in hilly and learningusingdeepvigorously even in difficult Situations such as drought, salinity, and waterlogging?“mountainous areas?“Aexampleofadeeplearningmodelthatcaneasilydetectpersimmonfruitinimagesandvideos.Itcanleadtoyieldpredictionandmoreefficientcultivationmanagement.Anexampleofamodelfordiscriminatingbetweensmall-seededfruitandlarge-seededfruit.Althoughthereisnodifferenceinappearancebetweenthetwofruits,deeplearningmodelsuggeststhatthefeaturesmaybepresentinthefruitapexandpeduncleside.Plant materials; millet and soybean Transverse cross section of adventitious root of Satariaglauca YutaroOSAKOAssistantProfessor, Ph. DAsana MatsuuraAssociate Professor,Ph.D.19Bioresource ScienceDivisionBioresource Science DivisionLaboratory of PomologyLaboratory of Crop Science
元のページ ../index.html#23