Skip to content

Latest commit

 

History

History
61 lines (54 loc) · 1.86 KB

README.md

File metadata and controls

61 lines (54 loc) · 1.86 KB
Raw

PBA vs G2O Times comparator

Compare the execution times of g2o for bundle adjustment and pba.

Getting started

Compile both project as standard C++ project.

Compile G2O

Open a terminal in current project direcory. Follow these instructions to compile g2o_ba:

cd g2o_ba/thirdparty/g2o/
mkdir build
cd build
cmake ..
make
cd ../../..
mkdir build
cd build
cmake ..
make

Compile PBA

Open a terminal in current project directory. Follow these instructions to compile pba:

cd pba/
mkdir build
cd build
cmake ..
make

Usage

Follow these instuctions to get the times in a .csvfile:

Untar input files

Untar the provided input files inside inputs/ directory. It the archive had size > 100MB we had splitted it in more parts. To join all the parts use this command:

cat archive_name.tar.bz2* > archive_name.tar.bz2

for example cat dubrovnik.tar.bz2* > dubrovnik.tar.bz2 will create the joined archive.

Launch the script

We provide a script which automatically load all the input files given a dataset and memorize its execution times in a .csv file inside the times/ directory. The script takes as input the dataset directory name inside inputs/ directory. For example to launch the script on the dubrovnik dataset

./gettimes.sh dubrovnik

When the execution is complete you will fild the files pba.csv and g2o.csv inside the times/ directory. Remember to rename these files before launch other simulations or the results will be appendend to these files.

Use your own dataset

You can use your own dataset just by:

  • Get the files in PBA format
  • Create a directory inside inputs/ directory
  • Move your file inside this directory
  • Launch the script cited above