This project is aimed to generate the datasets used in our paper "Geolocalization with aerial image sequence for UAVs". We generate aerial image sequences with the simulated flight of the UAVs with the cameras mounted to. For more details about how we do this, please refer to our article.
The lib is used to decode the mrsid file, where the color aerial ortho imageries are saved, we provide the related lib files in the directory thrid_lib.
The lib is used to decode the shp file, we provide the related lib files in the directory thrid_lib.
Change the path for opencv to your own path.
Install the Eigen3 with the instruction in http://eigen.tuxfamily.org/index.php?title=Main_Page.
Build the project with the following commands: $ cd ${root_path_of_the_project}
$ mkdir build
$ cd build
$ cmake ../
$ make
Download the color aerial ortho imageries from https://docs.digital.mass.gov/. Download the road vector map file from OpenStreetMap, and save it in the form of .shp.
Set the parameters in the python script in 'tools/make_image_sequences.py' according to your requirements.
Run the python script to generate the aerial image sequences using command:
$ python make_image_sequences.py