Presentation Information
[B-15-36]Quantitative analysis for transitional growth degree of crops applying machine learning
〇Yoshitsugu Nakagawa1, Hiroyasu Sano1, Toshiro Takata2 (1. Tokyo Metropolitan Industrial Technology Research Institute, 2. Nozomi Corp.)
Keywords:
Machine Learning,Light Spectral Distribution,Crop Growth Degree,Random Forest,Clustering
In recent years, Japanese agriculture has faced a serious shortage of successors, not only due to the aging of the farming population but also because the number of agricultural workers has halved in the past 20 years. This is not simply a matter of "smart agriculture" utilizing IT and AI technology being the solution; one contributing factor is that the artisanal "intuition and experience" and person-dependent work processes are not codified and are not being passed on to the next generation. In this research, we will analyze the differences in growth that experienced farmers describe in relation to the visual progression of crop growth by focusing on the changes in the light spectral distribution and using machine learning for clustering methods. This is expected to lead to data-driven agricultural management that visualizes and integrates environmental control and crop growth, guiding support for new farmers.