With the challenge of reducing pesticide use, early detection and monitoring of crop pests is essential in controlling pest dynamics. Colorado potato beetle (CPB) is one of the most devastating insect pests in potato crops and its impact on crop damage and subsequent yield loss is expected to increase even further due to climate change. This MSc thesis project aims to explore the feasibility of detecting the beetle and its abundance from the air by using data from high resolution cameras that were mounted on drones. A big data set is available with actual beetle counts on the potato plants and drone images for six dates over the last growing season (2020).
In this thesis project you will explore the feasibility of detecting and counting beetles from remote sensing images. The drone images have a resolution of a few mm and visually the beetles can be identified. But the challenge is how to identify them in all images for all dates. For this purpose we want to adopt advanced machine learning methods such as deep learning. This will also include the preparation of a training and validation dataset to develop the machine learning models.
This research is done in cooperation with the Crop Systems Analysis (CSA) group of Wageningen University.
- Prepare ortho-mosaics from irregularly acquired RGB photo’s with a UAV
- Prepare a training and validation dataset for CPB detection
- Develop a deep learning based approach for automated detection of CPB from very-high resolution RGB images
- Hunt, E.R. and Rondon, S.I., 2017. Detection of potato beetle damage using remote sensing from small unmanned aircraft systems. Journal of Applied Remote Sensing, 11(2), p.026013.
- Iost Filho, F.H., Heldens, W.B., Kong, Z. and de Lange, E.S., 2020. Drones: Innovative Technology for Use in Precision Pest Management. Journal of Economic Entomology, 113(1), pp.1-25.
- Roosjen, P.J. , Kellenberger, B. , Kooistra, L. , Green, D.R. , Fahrentrapp, J., 2020. Deep learning for automated detection of Drosophila suzukii : potential for UAV-based monitoring. Pest Management Science 76: 9.
- We are looking for a motivated student that is interested in this relevant topic, and likes to apply state of the art image analysis techniques on an important agricultural problem. The background knowledge needed to successfully complete this project are: data analysis skills, high-resolution drone remote sensing and experience with machine learning techniques.
Theme(s): Sensing & measuring; Integrated Land Monitoring