Due to the aging of craftsmen and the mechanization of agriculture, "Winnowing basket" as a kind of traditional farm tool, the handcraft of making it is gradually disappearing. This study aims to visualize and quantify physical tacit knowledge for making “winnowing basket” by a video with Openpose and 3D pose base line program, which are capable of extracting the pose of a person from deep learning as a different method than before. It would enable us to develop new approach to understand traditional techniques by utilizing the wide resource much more effectively, which is helpful in the transmission of the raditional craft education.