Cost volume filter

Summary

Implementation of the CostVolumeFilter application. It computes the disparity map for a pair of images in epipolar geometry.

Rationale

This application permits to adapt the method proposed by Rhemann and al, 2011, to large inputs.

Implementation Details

Classes and files

New filters have been added in order to :

• construct the disparity-space range
• extract the index of the minimum value in a volume
• compute the guided filter
• and fill the occlusions pixels.

Applications

The CostVolumeFilter application has been added to AppStereo folder.

Tests

Five tests have been developed for:

• the minimization of the cost volume filter
• the minimization of the guided filter
• the occlusions detection filter
• the densification process to fill the the disparity map
• finally, the test of the application allows to test all the filters implemented.

Documentation

Information has been added in the CookBook in the "Stereoscopic reconstruction from VHR optical images pair" part.

Additional notes

The CostVolumeFilter application works for RGB and grayscale images. A possible issue could be to adapt the code for N-bands image.

Copyright

The copyright owners are CS Systemes d'Information (CS SI) and Damia Belazouz, and they have signed the ORFEO ToolBox Contributor License Agreement.

---

***Check before merging:***

• All discussions are resolved
• At least 2 votes from core developers, no vote.
• The feature branch is (reasonably) up-to-date with the base branch
• Dashboard is green
• Copyright owner has signed the ORFEO ToolBox Contributor License Agreement