LeyangWen/stereo

Stereo calibration and disparity creation

This project is a basis for stereo vision. After calibration process, you can try to get disparity map and start your own project. You can check my blog if you are new to stereo vision. I wrote about what is stereo vision and how it works briefly.

Usage

Get images from single camera

Don't forget to create directory with mkdir and set your cameras in the code. Example usage:

python get_images.py directory 0

Get stereo images

Don't forget to set your cameras. Check if the right is right and left is left. Example usage:

python getStereoImages.py directory 0

Split concat images

This code is for fixed stereo set. There are some cameras that give you the most possible sync. They are taking the images and concat with hardware. You see that you have only one camera but actually you are getting two images as one. So we need seperation before proceed. Get images as single camera and run this code. Example usage:

python grabSeperate.py directory savedDirectory prefixOfImages

Single camera calibration

Example usage:

python single_camera_calibration.py data\leftFixedStereo --image_format png --prefix left --square_size 0.025 --width 9 --height 6 --save_file left_cam.yml

Stereo camera calibration

Example usage:

python stereo_camera_calibration.py --left_file left_cam.yml --right_file right_cam.yml --left_prefix left --right_prefix right --left_dir bothImagesFixedStereo --right_dir bothImagesFixedStereo --image_format png --square_size 0.025 --save_file stereo_cam.yml

Disparity map

Example usage:

python stereo_depth.py --calibration_file data\stereo_cam.yml --left_source data\left_1541067450.avi --right_source data\right_1541067450.avi --is_real_time 0