Flower Mapping in Grasslands With Drones and Deep Learning

Abstract

Manual assessment of flower abundance of different flowering plant species in grasslands is a time-consuming process. We present an automated approach to determine the flower abundance in grasslands from drone-based aerial images by using deep learning (Faster R-CNN) object detection approach, which was trained and evaluated on data from five flights at two sites. Our deep learning network was able to identify and classify individual flowers. The novel method allowed generating spatially explicit maps of flower abundance that met or exceeded the accuracy of the manual-count-data extrapolation method while being less labor intensive. The results were very good for some types of flowers, with precision and recall being close to or higher than 90%. Other flowers were detected poorly due to reasons such as lack of enough training data, appearance changes due to phenology, or flowers being too small to be reliably distinguishable on the aerial images. The method was able to give precise estimates of the abundance of many flowering plant species. In the future, the collection of more training data will allow better predictions for the flowers that are not well predicted yet. The developed pipeline can be applied to any sort of aerial object detection problem.

Data-centric AI approach for automated wildflower monitoring

“…Hicks et al [39], used it to derive a nectar sugar mass index from detected floral resources. And Gallman et al [60] adopted this AI technology along with drone-based image acquisition and the construction of georeferenced orthomosaics for detecting wildflowers in mountainous areas. The proposed solutions of Hicks et al [39] and Gallman et al [60] can identify and count 25 common wildflowers.…”

Unmanned aerial vehicles (UAV) utilization for mapping the health of oil palm plants (Elaeis guineensis Jacq.)

“…Studies on the identification and mapping of healthy plants that are attacked by BSR disease using images from multispectral cameras on UAVs are still rarely carried out [18,19]. Plant health conditions identification and the number of these plants using this drone can identify and classify oil palms, enabling the creation of more accurate oil palm population maps compared to manual calculations [20,21]. Based on this, research was carried out, namely the use of unmanned aerial vehicles for mapping the health of oil palm plants (Elaeis guineensis Jacq.).…”

Advances in high-throughput crop phenotyping using unmanned aerial vehicles (UAVs)

“…With the advent of deep learning, this approach has been largely automated. Today convolutional neural networks can be used to segment plants from soil even in challenging conditions (Zenkl et al, 'in review'), detect and even count flower heads (Gallmann et al, 2022) and wheat ears (David et al, 2020). The publications by David et al (2020David et al ( , 2021 are particularly interesting since they represent a worldwide initiative by many institutions to compile a dataset called the Global Wheat Head Dataset that now serves as a benchmark dataset for the global machine learning community.…”