/Generating-Disparity-Maps-From-Stereo-Images

using block matching and dynamic programming to generate disparity maps from stereo images.

Primary LanguageJupyter Notebook

generating-Disparity-maps-from-stereo-images

using block matching and dynamic programming to generate disparity maps from stereo images.

brief introduction

having 2 parallel stereo image we can deduce the depth from disparity. we can compute the depth using this equation depth = (baseline * focal length) / disparity)

using epipolar line, we match point X in 3D located in left image to the same point X right image and then calculate the disparity from the displacement between them as illustrated bellow.

example

here we can see two stereo images

right image

results

block matching (left) dynamic programming (right)

dynamic programming

extra resources and ref