Skip to main content
OpenConf small logo

Providing all your submission and review needs
Abstract and paper submission, peer-review, discussion, shepherding, program, proceedings, and much more

Worldwide & Multilingual
OpenConf has powered thousands of events and journals in over 100 countries and more than a dozen languages.

Classifying Forest Roads Using Geometric Features Detected With The Mobile Proximal Sensing Platform Roadsens

Sustainable forest management requires a well-planned and maintained forest road network to provide access to remote timber areas and for other management purposes. In Norway, most forest roads were built over 30 years ago, following partly outdated traffic requirements, and frequently being subjected to wear from transportation and deterioration of water-bound surfaces. To ensure traffic safety, mitigation of environmental impacts such as road-induced surface water channeling, and to avoid cost-intensive repairs, up-to-date information of the current technical condition and deterioration of the roads is essential. This information is often lacking and existing road databases are incomplete due to the main assessment methods conducted being manual, which are time-consuming and expensive. An alternative solution could be the use of remote sensing techniques such as Airborne Laser Scanning or aerial imagery. However, these alternatives have restrictions. Airborne Laser Scanning lacks frequent updates and the necessary point density for functional road classification, while aerial imagery, despite its frequent updates proves inefficient in forest areas with dense canopy coverage. Thus, proximal sensing technologies implemented through operational vehicle traffic on forest roads offer new opportunities to collect multiple road-related data in frequent updating intervals, at a much lower cost. The study presents an application of the proximal sensing platform RoadSens for technical road classification. The data products generated by RoadSens along with the 3D point cloud of the road include longitudinal and cross-sectional characteristics in a relatively high spatial resolution. Five different case study regions in Southern and Eastern Norway have been used to demonstrate the application of the platform. The extracted geometric features are integrated into a classification scheme, following the national standards for technical rural road classification (class 1-8), to determine the surveyed road’s technical standard. The classification will be done road-wise based on the applicable cross sections with detected road geometry to determine if the criteria are fulfilled, such as road width, ditch depth, and side-fall, set for the distinct technical class of the road standard. If sections of the road deviate from the standard it will be determined if the deviation leads to the entire road being degraded or if they can be accepted for the area. The goal of the study is to find a way to classify forest roads based on detected road geometry, as well as identify the roads most common limitations based on the national standards. This study is part of the Norwegian SmartForest project, where RoadSens is forming a component of the ForestSens cloud service family. Further developments of RoadSens will include an end-user application not requiring in-depth knowledge on data handling, processing, and interpretation. Where the output data will automatically show up as simple and readable results for the user in question. In summary, this study offers a novel approach to forest road classification that aims to overcome the limitations of traditional methods and remotely sensed techniques and provide a cost-effective and efficient solution to assess the technical condition of forest roads.

Helle Ross Gobakken
NIBIO, Department of Forest Operations and Digitalization
Norway

Mostafa Hoseini
NIBIO, Department of Forest Operations and Digitalization
Norway

Stephan Hoffmann
NIBIO, Department of Forest Operations and Digitalization
Norway

Jan Bjerketvedt
NIBIO, Department of Forest Operations and Digitalization
Norway

Johannes Rahlf
NIBIO, Department of National Forest Inventory
Norway

Rasmus Astrup
NIBIO, Division of Forest and Forest Resources
Norway