This repository contains a minimal code example to estimate disparity using rectified stereo images from the Middlebury Cones dataset. A local, fixed-support stereo correspondence algorithm is used. Specifically, a window-based method using the sum-of-absolute-differences (SAD) similarity measure is implemented.
Left Stereo Image | Right Stereo Image |
Ground Truth Disparity | Estimated Disparity |
- Create a conda environment and activate it.
conda create -n stereo python=3.9
conda activate stereo
- Clone the repository
git clone https://github.com/ed-winning/stereo_depth_estimation.git
- Install the dependencies
pip install -r requirements.txt
The following code will generate and visualize the estimated disparity image shown above:
python main.py