Inspiration
For my robotics team.
What it does
It takes a stereo camera's feed and maps the movement of the camera on a 2d plane.
How I built it
I built it using OpenCV's object detection code. I would apply object detection to two frames and track the features in the two frames. The code then calculates the disparity between the two features and estimates the movement of the camera through the disparity.
Challenges I ran into
The biggest challenges I ran into is first handling a large number of features to track and also tracking features across frames.
Accomplishments that I'm proud of
I am proud of learning open-cv from scratch. Also, I learned a lot about data processing and building projects in general as this was the first big project that I made.
What's next for Visual Odometry
So far, nothing is next as I have finished the code and it is working on Team 846 Robotics team's robot which placed second in a qualifying tournament and qualified for worlds.
Log in or sign up for Devpost to join the conversation.