Easy Plant Phenotyping by using Image processing and machine learning techniques
Speaker: Wei GUO, Ph.D, The University of Tokyo
Abstract: Developments of high-throughput plant phenotyping technologies have been dramatically accelerated by advanced image sensing and analyzing techniques. However, plant scientists still meet difficulties when they try to apply such technologies practically. For example, to train an object detection deep learning model requires manual annotations of a huge amount of images. In this webinar, Wei will introduce the lessons his team learned from various case studies and share the experience they gain. At last, Wei will also propose some of his ideas on how to expand the challenges with images for more practical data such as 3D point clouds or images acquired from field conditions, as well as how to provide easy-to-use solutions to the community.
CV: Wei GUO is currently working as an Assistant professor at The University of Tokyo, Japan. He participated in the establishment of "International Field Phenomics Research Laboratory", the first plant phenomics laboratory in Japan in 2017 as a core member. His researches focus on field-based phenotyping by using advanced sensing platforms and technologies such as drones and ground robots, image processing and machine learning approaches.