Host: The Japan Society of Mechanical Engineers
Name : [in Japanese]
Date : May 27, 2020 - May 30, 2020
This paper proposes to estimate inexpensively yield, grass height, nutrition CP (Crude Protein) and nutrition TDN (Total Digestible Nutrients) of pasture using visible light image. The reason for using visible light image acquired by a general camera is that the cost can be reduced compared to using a near infrared light camera. In this study, 205 points data collected by Aerial shooting and reaping survey was utilized. An estimation method proposed new formulas composed of only RGB which refer to conventional formulas such as NDVI, VARI and so on. Extracted RGB values from visible light images of pasture was assigned to RGB formulas. It was calculated correlation coefficient between targets of estimation and values of RGB formulas. As a result, it was suggested that estimating by proposed formulas of visible light is nearly equal to or higher accuracy than estimating by a conventional formula of near infrared light.
