Aerial Implicit 3D Video Stabilization Using Epipolar Geometry Constraint

Abstract

We present an accurate video stabilization method on aerial videos using the epipolar geometry constraint. Most previous methods used 2D homography for stabilization, but failed to overcome the parallax problem. In this work, we propose to use dense correspondences for stabilization and the epipolar constraint to deal with the parallax effect. We start by estimating the dense correspondences between two frames. The dense correspondences are then used to estimate the epipolar geometry. The epipolar geometry has an implicit 3D constraint that can be used to improve the dense correspondences and handle the parallax. We evaluate our method on a real-life database containing three aerial image sequences. We also compare our method with the dominant 2D method for aerial video stabilization. The quantitative result demonstrates the effectiveness of our approach.

Automated White Blood Cell Counting in Nailfold Capillary Using Deep Learning Segmentation and Video Stabilization

“…To minimize unwanted motion artifacts in a capillary video stream, the translation vectors of video frames corresponding to a reference frame are determined [ 20 ]. A motion artifact is measured between two frames; current frame and reference frame.…”

“…The magnification in capillary videos of small hand motions can produce such large camera movements, representing a potentially severe problem in the proposed method. A previous study [ 20 ] has addressed this problem by accumulating movements. Likewise, we adopt the same approach to reduce the extent of large movements.…”

“…Additionally, inaccurate capillary detection and miss-detection of the WBC event might occur as a small finger motion is magnified when seen through a microscope. To mitigate this type of artifact, we apply motion compensation while also adopting an efficient version of image registration for translations [ 19 ] that reduces the capillary motion and accumulates data in case of a large unwanted movement in the recorded capillary videos [ 20 ]. The main contribution of this study is the automation of WBC event counting using a deep learning approach supported by video stabilization for robustness against motion artifacts.…”