Click on “Download PDF” for the PDF version or on the title for the HTML version.
If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.
A Real-Time Route Planning Method Based on DeepLabV3+ for Plant Protection UAVs
Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan www.asabe.org
Citation: 2021 ASABE Annual International Virtual Meeting 2100018.(doi:10.13031/aim.202100018)
Authors: Junjie Wan, Lijun Qi, Hao Zhang, Zhong’ao Lu, Jiarui Zhou
Keywords: DeepLabV3+,MATLAB,Route planning,Semantic segmentation,Unmanned aerial vehicle
Abstract. At present, the flight path of plant protection drones is usually planned during operation, which is time-consuming, laborious, and poorly-timed, and cannot deal with sudden pests and diseases in orchards. To solve this problem, a real-time route planning method based on DeepLabV3+ for plant protection UAVs was proposed. Firstly, the DJI Phantom 3 aerial drone was used to obtain the image of the fruit tree canopy. Then, the DeepLabV3+ deep learning model is used to segment the image of the fruit tree canopy. Finally, the method of extracting navigation lines is automatically selected by judging the number of centroids of segmented binary images. The test results show that the accuracy of the algorithm for extracting navigation lines is 95%, and the average time to process a single image is about 0.55 s. The detection accuracy and real-time performance can meet the requirements of plant protection UAV operations. Besides, the method in this paper has good adaptability to the fracture of the fruit tree canopy.
(Download PDF) (Export to EndNotes)