Portugal has witnessed some of its largest wildfires in the last decade, due to the lack of forestry management and valuation strategies. A cost-effective biomass collection tool/approach can increase the forest valuing, being a tool to reduce fire risk in the forest. However, cost-effective forestry machinery/solutions are needed to harvest this biomass. Most of bigger operations in forests are already highly mechanized, but not the smaller operations. Mobile robotics know-how combined with new virtual reality and remote sensing techniques paved the way for a new robotics perspective regarding work machines in the forest. Navigation is still a challenge in a forest. There is a lot of information, trees consist of obstacles while lower vegetation may hide danger for robot trajectory, and the terrain in our region is mostly steep. The existence of accurate information about the environment is crucial for the navigation process and for biomass inventory. This paper presents a prototype forest robot for biomass collection. Besides, it is provided a dataset of different forest environments, containing data from different sensors such as 3D laser data, thermal camera, inertial units, GNSS, and RGB camera. These datasets are meant to provide information for the study of the forest terrain, allowing further development and research of navigation planning, biomass analysis, task planning, and information that professionals of this field may require.
Forest Robot and Datasets for Biomass Collection
Advs in Intelligent Syst., Computing
Iberian Robotics conference ; 2019 ; Porto, Portugal November 20, 2019 - November 22, 2019
2019-11-20
12 pages
Article/Chapter (Book)
Electronic Resource
English
Forest Robot and Datasets for Biomass Collection
TIBKAT | 2020
|Efficient Label Collection for Image Datasets via Hierarchical Clustering
British Library Online Contents | 2018
|British Library Conference Proceedings | 2014
|