논문 상세보기
Background : Due to changes in climate and cultivation conditions, the growth monitoring is an essential factor in improving crop productivity. With the recent development of image analysis technology incorporating ICT, it has become possible to constantly monitor the crop growth. As a medicinal crop specialized in Gyeongsangbuk-do, Korea, Cnidium officinale Makino was examined for the possibility of growth diagnosis through image analysis for stable production.
Methods and Results : The IP camera was installed at 2.5 m height in experiment field. The RGB image of every 06:00 was captured from July 1 to July 30 and used for analysis. The captured images were analyzed using the image analysis tool, Image J. The greeness was estimated by the average value of the green histogram. The canopy size was determined by the color range (red: 0-255, green: mean value-255, blue: 0-255) and was calculated as the ratio of pixels number of the entire image to those of the selected area. The growth temperature during investigation period was measured by Hobo MX2300. High temperature, excess of 28℃, was compared to stress response such as decrement of canopy size. The greeness and the canopy size are respectively represented by the quadratic function greeness = -0.0722GD2 + 6248.9GD – 1e + 08 (GD, growing day; R2 = 0.46) and canopy size = -0.0462GD2 + 3996.7GD – 9e + 07 (R2 = 0.93). From July 11, it began to exceed the growth limit temperature of 28℃, and the canopy size began to decrease from this period. Between the canopy size (C) and the accumulated temperature exceeding 28℃, there was a negative correlation, C = -0.13ATEC + 56.75 (R2 = 0.87) during the decreasing period.
Conclusion : Extraction of color information in Cnidium officinale Makino using RGB image should be preceded by standardized setting, but it is considered to be useful tool for analyzing the change of quantitative characteristics over time. In the future, it is necessary to make a comparative study with the actual growth rate in the image diagnosis.
