Fix ransac issue "SolvePnPRansac Fails because calls to solvePnP with less than 6 points" by Dylancer1998 · Pull Request #19492 · opencv/opencv

Dylancer1998 · 2021-02-10T04:36:03Z

Firstly, I proposed this issue in there. issue #17799
I did some amendment based on this PR.

My solutions:

return "true" and the EPnP's result when SolvePnPRansac Fails due to there is no coplanaer data and DLT with less than 6 points. Because I think the result of EPnP may be correct, we should leave the right for user to choose whether use or not use the result.
provide a test data
provide a log for user

Pull Request Readiness Checklist

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

I agree to contribute to the project under Apache 2 License.
To the best of my knowledge, the proposed patch is not based on a code under GPL or other license that is incompatible with OpenCV
The PR is proposed to proper branch
There is reference to 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

force_builders_only=linux,docs

Dylancer1998 · 2021-02-10T04:38:17Z

@catree Could you help me review this PR? Thanks!

alalek · 2021-02-10T09:25:21Z

~~It makes sense to start from the regression test.~~ Test is here

Dylancer1998 · 2021-02-10T10:52:51Z

~~It makes sense to start from the regression test.~~ Test is here

Is there any problem? Could you guide me what should I do?

catree

Quick review.

Probably we could try to fuse the two PRs.

catree · 2021-02-10T19:28:57Z

modules/calib3d/src/solvepnp.cpp

                      distCoeffs, rvec, tvec, useExtrinsicGuess,
                      (flags == SOLVEPNP_P3P || flags == SOLVEPNP_AP3P) ? SOLVEPNP_EPNP : flags) ? 1 : -1;
+    }
+    catch(...){


I think it is better to have catch (const cv::Exception& e).

And log also e.what(). This way, if another type of exception is thrown from solvePnP it will be logged.

Ok. I agree with you. I have already commited the modification.

catree · 2021-02-10T19:33:41Z

modules/calib3d/test/test_solvepnp_ransac.cpp

+    camera_mat.at<float>(1, 2) = 353.712;
+    camera_mat.at<float>(2, 2) = 1.f;
+
+    pts3ds = {


/build/precommit_linux64/3.4/opencv/modules/calib3d/test/test_solvepnp_ransac.cpp:2247:5: warning: extended initializer lists only available with -std=c++11 or -std=gnu++11 [enabled by default]
/build/precommit_linux64/3.4/opencv/modules/calib3d/test/test_solvepnp_ransac.cpp:2238:12: warning: extended initializer lists only available with -std=c++11 or -std=gnu++11 [enabled by default]
/build/precommit_linux64/3.4/opencv/modules/calib3d/test/test_solvepnp_ransac.cpp:2238:12: warning: extended initializer lists only available with -std=c++11 or -std=gnu++11 [enabled by default]
/build/precommit_linux64/3.4/opencv/modules/calib3d/test/test_solvepnp_ransac.cpp:2238:12: warning: extended initializer lists only available with -std=c++11 or -std=gnu++11 [enabled by default]
/build/precommit_linux64/3.4/opencv/modules/calib3d/test/test_solvepnp_ransac.cpp:2258:5: warning: extended initializer lists only available with -std=c++11 or -std=gnu++11 [enabled by default]
/build/precommit_linux64/3.4/opencv/modules/calib3d/test/test_solvepnp_ransac.cpp:2249:12: warning: extended initializer lists only available with -std=c++11 or -std=gnu++11 [enabled by default]
/build/precommit_linux64/3.4/opencv/modules/calib3d/test/test_solvepnp_ransac.cpp:2249:12: warning: extended initializer lists only available with -std=c++11 or -std=gnu++11 [enabled by default]
/build/precommit_linux64/3.4/opencv/modules/calib3d/test/test_solvepnp_ransac.cpp:2249:12: warning: extended initializer lists only available with -std=c++11 or -std=gnu++11 [enabled by default]

See the buildbot (you can click on the build numbers): https://pullrequest.opencv.org/#/summary/opencv

This requires C++11 and 3.4 OpenCV cannot require C++11 features.

Please help me what should I do to fix the build failures?

Initializer list for vector is a C++11 feature which should not be used here. You can either push_back the points to the vector, or use cv::Mat instead. See my PR here.

mightbxg · 2021-02-11T06:11:22Z

modules/calib3d/src/solvepnp.cpp

+    catch(const cv::Exception& e)
+    {
+        CV_LOG_DEBUG(e.what());
+        CV_LOG_DEBUG(NULL, "Broken implementation for DLT algorithm needs at least 6 points for pose estimation. Fallback to EPnP.");


Log message here is inaccurate. You cannot just presume that the Exception can only be caused by DLT algorithm.

mightbxg · 2021-02-11T06:17:02Z

modules/calib3d/src/solvepnp.cpp

+        if(_inliers.needed())
+        {
+            Mat _local_inliers;
+            for (int i = 0; i < npoints; ++i)
+            {
+                if((int)_mask_local_inliers.at<uchar>(i) != 0) // inliers mask
+                    _local_inliers.push_back(i);    // output inliers vector
+            }
+            _local_inliers.copyTo(_inliers);
+        }
+
+        return true;


Unnecessary codes (L329-L340)

mightbxg · 2021-02-11T06:21:41Z

@Johnzdh
This PR seems just a duplicate of mine #19253

Dylancer1998 · 2021-02-11T12:12:38Z

@Johnzdh
This PR seems just a duplicate of mine #19253

Great job~ Thanks!
I saw your previous PR was to return false when triggered the DLT error, so I wanted to fixed it.
I agree with your later PR. When your later PR is merged, I will close this one.

fix-ransac-issue

5e9543b

mightbxg mentioned this pull request Feb 10, 2021

Unexpected exception in solvePnPRansac caused by input points #19253

Merged

6 tasks

catree reviewed Feb 10, 2021

View reviewed changes

fix-ransac-issue

a022501

mightbxg reviewed Feb 11, 2021

View reviewed changes

Uh oh!

Conversation

Dylancer1998 commented Feb 10, 2021 • edited by alalek Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

My solutions:

Pull Request Readiness Checklist

Uh oh!

Dylancer1998 commented Feb 10, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

alalek commented Feb 10, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Dylancer1998 commented Feb 10, 2021

Uh oh!

catree left a comment

Choose a reason for hiding this comment

Uh oh!

catree Feb 10, 2021

Choose a reason for hiding this comment

Uh oh!

Dylancer1998 Feb 11, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

catree Feb 10, 2021

Choose a reason for hiding this comment

Uh oh!

Dylancer1998 Feb 11, 2021

Choose a reason for hiding this comment

Uh oh!

mightbxg Feb 11, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

mightbxg Feb 11, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

mightbxg Feb 11, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

mightbxg commented Feb 11, 2021

Uh oh!

Dylancer1998 commented Feb 11, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Dylancer1998 commented Feb 10, 2021 •

edited by alalek

Loading

Dylancer1998 commented Feb 10, 2021 •

edited

Loading

alalek commented Feb 10, 2021 •

edited

Loading

Dylancer1998 Feb 11, 2021 •

edited

Loading

mightbxg Feb 11, 2021 •

edited

Loading

mightbxg Feb 11, 2021 •

edited

Loading

mightbxg Feb 11, 2021 •

edited

Loading

Dylancer1998 commented Feb 11, 2021 •

edited

Loading