• Intelligent Agriculture towards Big Data - Advanced Airborne LiDAR Technology Applied to Fruit Tree DetectionPest Control

  • 2019 -10 -31
Airborne LiDAR has the capability to survey a large area at high accuracy within a short period of time. Compared to traditional optical aerial photography and satellite imagery, airborne LiDAR utilizes laser ranging method to obtain the terrain and elevation data. This advanced sensing technology is suitable for estimating the height, width and volume of various fruit trees. Furthermore, the airborne LiDAR is also appropriate for numerous intelligent agricultural topics such as crop growth monitoring, fruit tree species classification, fruit yield prediction, and pest control.
Our team acquired a total number of 49.2 billion measurement data by airborne LiDAR scanning which covers the central and south Taiwan. The H-max method was chosen among other algorithms as the most suitable method for fruit tree detection which gives an accuracy of tree position higher than 80% and less than 15 cm tree height error. The results showed that there were 1.84 million Lychee trees in Taiwan with an average tree height of 4.67 m; 1.86 million Citrus trees with an average tree height of 2.72 m; and 0.28 million Wax-apple trees with an average height of 2.74 m.
The application of this advanced sensing technology in three intelligent agriculture topics as following:
(1) Fruit yield estimation. LiDAR data is utilized to estimate the yield of various fruits in each township.
(2) Fruit pest control. The tree height is used to distinguish the abandoned orchard or poorly growing fruit trees for prevention and control of the Lychee giant stink bugs (Tessaratoma papillosa), and for hot spots area detection of the Yellow Dragon disease (Huang-Long bing).
(3) Fruit tree detection in the shade-net house. The LiDAR’s laser light can penetrate the shade-net house to reach the fruit trees inside. Therefore, the number, height and species of the trees inside could be obtained by LiDAR scanning which cannot be achieved by traditional sensing technology.
Our team successfully introduced an advanced remote sensing technology for intelligent agriculture topics, and effectively processed the big data with high-accuracy result in short-period of time.