Could anybody suggest an automatic way to convert from a list of images (without Kinect) to a point cloud in opencv?
Asked
Active
Viewed 2,457 times
1 Answers
1
Take a look at OpenCV Contrib Structure From Motion module (SFM). There are two nice examples trajectory_reconstruction.cpp
and scene_reconstruction.cpp
.
Also, there is alternative called Multi-View Environment which you could find on GitHub at simonfuhrmann/mve and which might meet your criteria too.

Rasim
- 1,276
- 1
- 11
- 24