Disparity and depth map generation
Monoscopic video to 3D

Take a look at this article on the new site! Follow the link

Project: Dr. Dmitriy Vatolin
Implementation, algorithm: Sergey Grishin,
Karen Simonyan,
Alexander Parshin,
Konstantin Strelnikov

This technology provides high quality depth map estimation for video. Our algorithm doesn't require stereo content. Only information from single view video is used. Algorithm estimates motion in video sequence, generates disparity map and depth map. Similar technology can be used to convert stereo video to multiview 3D video.


Algorithm uses spatial and temporal information from consecutive frames of monoscopic video

Previous frame Current frame
Previous frame Current frame

to estimate disparity map or depth map:

Depth map anв current frame
Depth map vs current frame

Also stereo pair or multiview 3D video can be generated using the depth map:

Stereo pair
Stereo pair

Higher quality of multiview 3D video and depth/disparity map is achieved when stereoscopic material is used as the source of conversion.

More examples

Previous frame Current frame
Previous frame Current frame
Depth map Stereo pair
Generated depth map vs current frame Generated stereo pair


For questions and proposition please contact us  

Video resources:

Last updated: 12-May-2022

Server size: 8069 files, 1215Mb (Server statistics)

Project updated by
Server Team and MSU Video Group

Project sponsored by YUVsoft Corp.

Project supported by MSU Graphics & Media Lab