Michael Jae-Yoon Chung [email protected]
Initially, I wanted to use my assignment3 as a starting point of the final project. I wanted to make a web-interface that is capable of displaying on-board sensor data in real-time. However, I wasn't sure about the main purpose of such visualization tool. After couple of brainstorming sessions, I concluded that the visualization tool that supports sensor data inspection and comparison would be very useful. So I started to analyze widely used tools such as rviz and rqt_bag, and tried to improve it to better support sensor data inspection and comparason.
My code is heavily relying on ROS. First, you will need to install ROS Hydro on your machine. Theoratically, ROS can be installed on various platforms, however, it works the best with Ubuntu platform. Therefore, I only have tested my code using Ubuntu machines.
Once ROS is installed, you will need to setup your ROS environment as shown here. You can now run follow commands from a terminal to build rqt_bag_diff and launch roscore.
git clone https://github.com/CSE512-14W/fp-mjyc
cp fp-mjyc/rqt_bag_diff ~/catkin_ws/src/
cd ~/catkin_ws/
catkin_make
roscore
Finally run below command to launch rqt_bag_diff from another terminal.
python ~/catkin_ws/src/rqt_bag_diff/scripts/rqt_bag_diff
You can load any .bag formatted data you want to analyze using the load GUI button on the upper-left corner of the interface.
Please see the finals paper for the detailed usage explanations.
- Due to the heck I used (see bag_widget.py), you need to have "some_sensor_data.bag" file as well as the copy of that file with filename "some_sensor_data (copy).bag".
- The biggest limitation of rqt_bag_diff is how the program is currently managing "time". Timestamps from two data sources can be different, this can be problematic for ROS based viewers such as rviz and rqt_plot.