angubenko's profile - activity

overview network karma followed questions activity

2017-08-13 07:50:06 -0600	received badge	● Student (source)
2015-12-04 23:34:44 -0600	received badge	● Enthusiast
2015-12-01 03:03:40 -0600	received badge	● Scholar (source)
2015-11-30 03:47:53 -0600	commented answer	OpenCV Why does number of SIFT keypoints go over threshold? Thank you. This part "The number of best features to retain" of description seems a little bit confusing. Why do we need these extra keypoints if they have same rank? What usage of these extra keypoints? Do you think I should open an issue?
2015-11-30 03:43:48 -0600	commented question	OpenCV Why does number of SIFT keypoints go over threshold? That's what I did, thanks.
2015-11-30 03:42:35 -0600	received badge	● Supporter (source)
2015-11-29 15:20:43 -0600	commented question	OpenCV Why does number of SIFT keypoints go over threshold? That sounds like a bug to me. In this case algorithm should just ignore other features. Accorfing documentation it supposes to keep numer of best features.
2015-11-29 01:38:47 -0600	asked a question	OpenCV Why does number of SIFT keypoints go over threshold? Hi, I created SIFT constructor, with Threshold = 100 cv::Ptr<feature2d> f2d = xfeatures2d::SIFT::create(100); f2d->compute(image, keypoint, descriptor); Then I extract descriptors from sequence of images and for some images I receive number of keypoints bigger than limit by one. Please see descriptors print below: [128 x 101] [128 x 100] [128 x 100] [128 x 101] [128 x 100] [128 x 100] [128 x 101] [128 x 100] Why does it happen? Do have incorrect understanding of nfeatures attribute?