Final project for CSC420 introduction to computer vision course at the University of Toronto
- All of the code can be run using
main.py
- Browse for available images in
/image_pairs.
- The first argument for running main.py is the image name prefix, followed by a number.
- The number can be only
1
or2
, where1
indicates running the Anaglyph generation code, and2
runs the Disparity Map Generation. - Example:
python main.py veronica 1
will run Anaglyph code for theveronica_left
andveronica_right
image pair. - Example:
python main.py foot 2
will run Stereo Disparity code for thefoot_left
andfoot_right
image pair. There are more prompts following this initial command, that ask to run the Middlebury evaluation vs. StereoBM computation, and the Gaussian kernel size.
- Before running the Middlebury Evaluation, make sure that
imagemagick
is installed. - Video and blender files of final results using disparity are in 3D_model_results.
Enjoy! :)