Weed control is one of the biggest challenges in organic farms or nature reserve areas where mass spraying is prohibited. Recent advancements in remote sensing and airborne technologies provide a fast and efficient means to support environmental monitoring and management, allowing early detection of invasive species. However, in order to perform weed classification, current studies mostly relied on object-based image analysis (OBIA) and proprietary software which required substantial human inputs. This paper proposes an open-source workflow for automated weed mapping using a commercially available unmanned aerial vehicle (UAV). The UAV was flown at a low altitude between 10 m and 20 m, and collected truecolour RGB imagery over a weed-infested nature reserve. The aim of this study is to develop a repeatable and robust system for early weed detection, with minimum human intervention, for classification of Rumex obtusifolius (R. obtusifolius). Preliminary results of the proposed workflow achieved an overall accuracy of 92.1% with an F1 score of 78.7%. The approach also demonstrated the capability to map R. obtusifolius in datasets collected at various flight altitudes, camera settings and light conditions. This shows the potential to perform semi-or fully automated early weed detection system in grasslands using UAV-imagery.
Rumex obtusifolius (R. obtusifolius) is one of the most common non-cultivated weed in European grasslands. Its broad-leaved and wide-spread nature make this weed competitive with the native pasture species reducing grass yield (van Evert et al. 2010), while its oxalic acid content makes this species poisonous for livestock if large doses are consumed (Hejduk and Doležal 2004). Therefore, early removal is preferred especially in organic dairy farms or conservation areas where mass spraying is prohibited. Remote sensing and airborne technologies offer fast and efficient support in environmental monitoring allowing early detection of invasive species, yet current studies mostly rely on object-based image analysis (OBIA) and proprietary software to perform weed classification that require substantial human inputs. In this work, an open source workflow for automatic weed detection using unmanned aerial vehicle (UAV) RGB-imagery of native grassland had been developed using deep learning techniques, based on a previously developed OBIA approach (Lam et al. 2019). During the study, DJI Phantom 3 and 4 Pro were used for data acquisition throughout the vegetation period in 2018 and early 2019 at a nature conversation area in North Rhine-Westphalia, Germany. Images were processed using OpenDroneMap to produce orthomosaics. OBIA methods were then performed using Python and QGIS to assist the data labelling process for training a convolutional neural network (CNN), which was later used as an image classifier. Preliminary results of the proposed workflow achieved an overall accuracy of 93.8% and had demonstrated the capability in mapping R. obtusifolius in datasets collected at various flight altitudes, camera settings and light conditions. This shows the potential of developing a repeatable and robust system for semi-or fully-automated early weed detection in grassland using UAV-imagery.
The acquisition, storage, and processing of huge amounts of data and their fast analysis to generate information is not a new approach, but it becomes challenging through smart decision-making on the choice of hardware and software improvements. In the specific cases of environment protection, nature conservation, and precision farming, where fast and accurate reactions are required, drone technologies with imaging sensors are of interest in many research groups. However, post-processing of the images acquired by drone-based sensors such as the generation of orthomosaics from aerial images and superimposing the orthomosaics on a global map to identify the exact locations of the interested area is computationally intensive and sometimes takes hours or even days to achieve desired results. Initial tests have shown that photogrammetry software takes less time to generate an orthomosaic by running them on a workstation with higher CPU, RAM and GPU configurations. Tasks like setting up the application environment with dependencies, making this setup portable and manage installed services can be challenging, especially for small-and-medium-sized enterprises that have limited resources in exploring different architectures. To enhance the competitiveness of the small and medium-sized enterprises and research institutions, the accessibility of the proposed solution includes the integration of open-source tools and frameworks such as Kubernetes (version v1.13.4, available online: https://kubernetes.io/) and OpenDroneMap (version 0.3, available online: https://github.com/OpenDroneMap/ODM) enabling a reference architecture that is as vendor-neutral as possible. Current work is based on an on-premise cluster computing approach for fast and efficient photogrammetry process using open source software such as OpenDroneMap combined with light-weight containerization techniques such as Docker (version 17.12.1, available online: https://www.docker.io/), orchestrated by Kubernetes. The services provided by OpenDroneMap enable microservice-based architecture. These container-based services can be administered easily by a container orchestrator like Kubernetes. After setting up the servers with core OpenDroneMap services on our container-based cluster with Kubernetes as the orchestrator engine, the plan is to use the advantages of Kubernetes' powerful management capabilities to help maximize resource efficiency as the basis for creating Service Level Agreements to provide a cloud service.
A high-resolution UAV-borne LiDAR system with a Velodyne VLP16-Lite at its core was developed for surveying applications. The LiDAR unit was combined with a high-end IMU-GNSS solution for direct georeferencing (APX-15) and a single-board computer for data acquisition (2nd-gen. Intel NUC). Hardware and software solutions were developed for system integration. Moreover, a mechanical mount for isolating the sensitive components of the system from the UAV’s high-frequency vibration was built and evaluated. System architecture and preliminary results were presented. Furthermore, a sensitivity analysis revealed the system’s most important sources of error and suggested ways to overcome these.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.