A5083552558- Smart Agriculture and AI
- Remote Sensing in Agriculture
- Biological Control of Invasive Species
- Remote Sensing and LiDAR Applications
- Species Distribution and Climate Change
- Wood and Agarwood Research
Aarhus University
2019-2020
For decades, significant effort has been put into the development of plant detection and classification algorithms. However, it difficult to compare performance different algorithms, due lack a common testbed, such as public available annotated reference dataset. In this paper, we present Open Plant Phenotype Database (OPPD), dataset for classification. The contains 7590 RGB images 47 species. Each species is cultivated under three growth conditions, provide high degree diversity in terms...
Lack of annotated data for training deep learning systems is a challenge many visual recognition tasks. This especially true domain-specific applications, such as plant detection and recognition, where the annotation process can be both time-consuming error-prone. Generative models used to alleviate this issue by producing artificial that mimic properties real data. work presents semi-supervised generative adversarial network (GAN) model produce samples seedlings. By applying approach, we...
Recent advances in the Unmanned Aerial System (UAS) safety and perception systems enable safe low altitude autonomous terrain following flights recently demonstrated by consumer DJI Mavic PRO Phamtom 4 Pro drones. This paper presents first prototype system utilizing this functionality form of semi-automated UAS based collection crop/weed images where embedded ensures a significantly safer faster gathering weed with sub-millimeter resolution. The is to be used when weeds are at cotyledon...
Recent advances in the Unmanned Aerial System (UAS) safety and perception systems enable safe low altitude autonomous terrain following flights recently demonstrated by consumer DJI Mavic PRO Phamtom 4 Pro drones. This paper presents first prototype system utilizing this functionality form of semi-automated UAS based collection crop/weed images where embedded ensures a significantly safer faster gathering weed with sub-millimeter resolution. The is to be used when weeds are at cotyledon...