potok/opencv
1
0
Fork 0
mirror of https://github.com/opencv/opencv.git synced 2026-07-31 00:03:03 +04:00
Code Packages Releases Wiki Activity

Merge pull request #26221 from asmorkalov:as/refactor_multiview_interface

Browse Source 
Reworked multiview calibration interface #26221

- Use InputArray / OutputArray
- Use enum for camera type
- Sort parameters according guidelines
- Made more outputs optional
- Introduce flags and added tests for intrinsics and extrinsics guess.

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [ ] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
This commit is contained in:
Alexander Smorkalov
2024-10-01 16:53:16 +03:00
committed by GitHub
parent 6c743f8e4b
commit 3bcab8db0a
8 changed files with 362 additions and 214 deletions
Download Patch File Download Diff File Expand all files Collapse all files
doc/tutorials/calib3d/camera_multiview_calibration/multiview_calibration.markdown
+1 -1
View File
@@ -162,7 +162,7 @@ Details Of The Algorithm
3. Output of intrinsic calibration also includes rotation, translation vectors (transform of pattern points to camera frame), and errors per frame. For each frame, the index of the camera with the lowest error among all cameras is saved.
2. Otherwise, if intrinsics are known, then the proposed algorithm runs perspective-n-point estimation (@ref cv::solvePnP, @ref cv::fisheye::solvePnP) to estimate rotation and translation vectors, and reprojection error for each frame.
2. **Initialization of relative camera pose**.
1. If the initial relative poses are not assumed known (the `useExtrinsicsGuess` is set false), then the relative camera extrinsics are found by traversing a spanning tree and estimating pairwise relative camera pose.
1. If the initial relative poses are not assumed known (CALIB_USE_EXTRINSIC_GUESS flag not set), then the relative camera extrinsics are found by traversing a spanning tree and estimating pairwise relative camera pose.
1. **Miminal spanning tree establishment**. Assume that cameras can be represented as nodes of a connected graph. An edge between two cameras is created if there is any concurrent observation over all frames. If the graph does not connect all cameras (i.e., exists a camera that has no overlap with other cameras) then calibration is not possible. Otherwise, the next step consists of finding the [maximum spanning tree](https://en.wikipedia.org/wiki/Minimum_spanning_tree) (MST) of this graph. The MST captures all the best pairwise camera connections. The weight of edges across all frames is a weighted combination of multiple factors:
* (Major) The number of pattern points detected in both images (cameras)
* Ratio of area of convex hull of projected points in the image to the image resolution.