Off-Road Vehicle Trafficability Maps Developed via Remote Sensing and Machine Learning

A methodological framework for generating vehicle trafficability maps using spatial modeling techniques. The model integrates biophysical, infrastructure, and meteorological variables via remote sensing, Geographic Object-Based Image Analysis (GEOBIA), Random Forest classification, and Multicriteria Decision Analysis (MCDA). The work was authored by Renan Fabres Dalmonech and last updated on 2026-05-11.

Use Cases

• Transportation planning based on integrated spatial layers for relief, land use, and soil.
• Environmental management based on biophysical and meteorological variables like wind speed and precipitation.
• Military operations planning based on a continuous Trafficability Index and categorical mapping.
• Cartographic modeling research based on the formalized, generalizable procedure for terrain analysis.

Strengths

• Model validation achieved an overall accuracy of 84.7% and a Kappa index of 0.81 via field surveys.
• Methodology integrates heterogeneous spatial layers (e.g., relief, land cover, soil, wind speed, precipitation, temperature) and non-spatial data (e.g., vehicle tire type).
• Framework is described as replicable and generalizable for diverse applications.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset is small (66.5 KB), suggesting limited scope, likely containing documentation rather than extensive raw data.

Provenance

Source

figshare

Collection Method

Developed via refined spatial modeling integrating remote sensing data, GEOBIA, Machine Learning (Random Forest), and Multicriteria Decision Analysis (AHP) within the Dinamica EGO platform.

Freshness

Last updated 2026-05-11 09:06:33; freshness should be verified.