< I am trying to run opencv 3.3 multicamera calibration on a set of RGB images captured by a RGBD sensor. When I run on matlab I get many features as below:
Camera #2: Adding photo #98...
....Matches: 1473
....Matches after Homog. Check: 62
....Matches after smoothness Check: 58
....58 features kept
When I use the opencv multicamera calibration code i get a constant of 4 features filtered for every image:
open image 0-89.jpg successfully
number of matched points 163
number of filtered points 4
...
I am attempting to use the multicamera calibration library and I am getting the same exact problem as you (only keeps 4 points from 1200+ matched points). Were you able to solve this problem? If so, how?
System information (version)
Detailed description
< I am trying to run opencv 3.3 multicamera calibration on a set of RGB images captured by a RGBD sensor. When I run on matlab I get many features as below: Camera #2: Adding photo #98... ....Matches: 1473 ....Matches after Homog. Check: 62 ....Matches after smoothness Check: 58 ....58 features kept
When I use the opencv multicamera calibration code i get a constant of 4 features filtered for every image:
open image 0-89.jpg successfully number of matched points 163 number of filtered points 4 ...
Steps to reproduce
< opencv-3.3.0/build/bin/example_ccalib_multi_cameras_calibration -nc 2 -pw 21211 -ph 15000 -ct 0 -fe 0 -nm 15 -v 1 imagelists.yaml -->