Skip to content

Algorithms and dataset developed in our research about strategies and evaluation methods for ground global path planning based on aerial images.

License

Notifications You must be signed in to change notification settings

cosmonautd/Aerial

Repository files navigation

Aerial Algorithms & Aerial Traversability and Planning Dataset

A strategy and evaluation method for ground global path planning based on aerial images

This repository contains the algorithms and dataset developed in our research about strategies and evaluation methods for ground global path planning based on aerial images.

If you use this repository, please cite our paper.

Algorithms

The module trav does traversability calculation, while graphmapx generates traversability graphs and paths. The examples script contains small snippets that show how to use the aforementioned modules. The experiments script contains the code required to reproduce the experiments in our paper.

Traversability and Planning Dataset

The Aerial Traversability and Planning Dataset is in the dataset directory. Following is a brief description of its contents.

  • images: aerial RGB images with resolution 1000x1000 pixels and approx. 0.3 x 0.3 m²/p;
  • labels: binary traversability labels for each sample in images;
  • keypoints-reachable: marked points of interest to use as source and target points;
  • keypoints-unreachable: marked points to test whether a path planning approach is able to identify infeasible paths.

Path Planning Samples

Sample 1 Sample 2

Sample 3 Sample 4

Sample 5 Sample 5

Sample 6 Sample 6

Sample 7 Sample 8

About

Algorithms and dataset developed in our research about strategies and evaluation methods for ground global path planning based on aerial images.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Languages