/stereo-vision-and-segmentation

Disparity Maps and Image Segmentation Implementation

Primary LanguagePython

CSE573 - Computer Vision and Image Processing - PA2

Stereo Vision and Image Segmentation

The code implements the requirements mentioned in PA2.pdf

This implements 1.1 and 1.2 from the assignment. Disparity Maps are generated from provided images using block sizes of 3 and 9. Back-projection is used to do consistency checks with the given ground-truth. Mean squared error for all the cases is also calculated.

Disparity Maps are generated using Dynamic Programming algorithm provided at section 4.1 in reference.

view3 is generated using the provided images and their ground-truths.

Mean Shift Segmentation is implemented and run on the given Butterfly image.