CameraRigBA::matchFrameToFrame allows to use only SURF_GPU for feature detection and matching

[sandeepnmenon]

As shown in the code below

camodocal/src/calib/CameraRigBA.cc

Lines 1476 to 1483 in b399fa9

	cv::Ptr<SurfGPU> surf = SurfGPU::instance(200.0);
	std::vector<cv::KeyPoint> rkeypoints1, rkeypoints2;
	cv::Mat rdtors1, rdtors2;
	std::vector<cv::DMatch> rmatches;
	surf->match(rimg1, rkeypoints1, rdtors1, rmask1,
	rimg2, rkeypoints2, rdtors2, rmask2, rmatches);

Unlike in FeatureTracker.cc, there is no support for switching between detectors and matchers.
The code can be improved here to accept m_detectorType, m_descriptorType and m_matchTestType

Are PRs accepted? I see no PRs accepted post 2017

[hengli]

PRs are definitely accepted. Due to lack of time, I have not been maintaining this repository, and help with this would definitely be appreciated.

…

On Thu, 26 May 2022 at 8:23 AM, Sandeep N Menon ***@***.***> wrote: As shown in the code below https://github.com/hengli/camodocal/blob/b399fa9ee81b5182e59f717e67a8ddd6503cdcb4/src/calib/CameraRigBA.cc#L1476-L1483 Unlike in FeatureTracker.cc, there is no support for switching between detectors and matchers. The code can be improved here to accept m_detectorType, m_descriptorType and m_matchTestType Are PRs accepted? I see no PRs accepted post 2017 — Reply to this email directly, view it on GitHub <#117>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AADXAZWXHAHNT4OPZU47QM3VL5ULVANCNFSM5XBAYSIQ> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>