/multicamera3DMotionReconstruction

3D position reconstruction with pyton and openCV

Primary LanguagePythonMIT LicenseMIT

MulticameraTraking

That's my bachelor thesis in coputer engeneering. It reconstructs 3d position of objects viewed by more than one camera. Object detection is implemented with a simple hsv color based method and 3d point estimation with a least square method. It's all documented on the thesis.pdf file.

Result

Some result can be seen on this animations. The first one is the ball bounce. This is the video of a bouncing ball captured by one of the two cameras and his reconstruction:

This is me trying to test better the depth and of course also the 3d reconstruction: