Study on Fire Blight Forecasting Using Rotary-Wing Drone

Hyun-Jung Kim1,*   Jun-Woo Park2   Chan-Seok  Ryu22   Tae-Hwan  Kang3,*   

1Department of Biosystems Engineering, Chungbuk National University, Cheongju, Republic of Korea
2Department of Bio-System Engineering, Gyeongsang National University, Jinju, 52828, Republic of Korea
3Major in Bio-Industry Mechanical Engineering, Kongju National University, Yesan, Republic of Korea

Abstract

The fire-blight disease is one of the contagious diseases infecting apples, pears, and some other species belonging to the Rosaceae family. Due to its extremely strong infectivity, once the infection is confirmed in the orchard, all of the tress located within 100 m must be destroyed and then the orchard is prohibited to cultivate the Rosaceae trees for 5 years. In South Korea, since fire-blight was confirmed for the first time in Ansung area in 2015, the infection is still being detected every year. Traditional methods to observe fire-blight requires a lot of costs and time, and also the pathogen can be transmitted by the inspectors from the infected orchard to the normal One. Therefore, it is necessary to develop a remote, unmanned monitoring system for fire-blight to prevent the communication of fire-blight. The pear tree image data were obtained at some pear orchards in Cheonan, using a multi-spectral camera with 4 spectral bands (red, green, blue, red-edge, and NIR) attached on a rotary-wing drone (DJI Phantom 4 PRO V2.0). The flight altitude was 15 m. Using Pix4D software, the still-cut images were connected into a single image and the spectral reflection values of the infected and normal tress were calculated from each band. As a result, the NIR band showed the most potential to identify the infected area in the orchard.

Figures & Tables

Fig. 1. Autonomous flight path in Cheonan