Image-based Phenotyping-from Images to Parameters to Information
Addressing the challenges of climate change and global food security,image-assisted phenotyping plays a major role in future plant breeding.Here,we introduce the field phenotyping facilities at the chair of plant breeding at University of Bonn.Based on images of 12 winter wheat varieties and captured with standard RGB cameras under field conditions,different phenotypic parameters were developed.An automated image processing pipeline using image analysis methods and machine learning approaches has been set up to compute seedling counts,plant ground cover and plant tissue vitality.Measuring large areas within plots,we improve phenotyping precision and representativity and accelerate phenotyping throughput.The ability to import,store and visualize data and corresponding images within our new crop information system enables breeders to directly benefit from image-based phenotyping.
phenotyping wheat image processing machine learning information system
Andreas Honecker Henrik Schumann Diana Becirevic Lasse Klingbeil Kai Volland Steffi Forberig Hinrich Paulsen Heiner Kuhlmann and Jens Léon
INRES-Plant Breeding,University of Bonn,Katzenburgweg 5,53115 Bonn,Germany IGG-Geodesy,University of Bonn,Nussallee17,53115 Bonn,Germany terrestris GmbH&Co.KG,K?lnstraβe 99,53111 Bonn,Germany
国内会议
北京
英文
199-208
2019-10-19（万方平台首次上网日期，不代表论文的发表时间）