Blur Robust Optical Flow using Motion Channel

TL;DR

This paper introduces a hybrid method combining camera motion data with optical flow estimation to improve accuracy in videos affected by motion blur, outperforming existing methods on both synthetic and real sequences.

Contribution

It proposes a novel hybrid framework that integrates camera motion information into optical flow estimation to handle motion blur more effectively.

Findings

Improved optical flow accuracy on blurry sequences.

Effective integration of camera motion data into the optical flow pipeline.

Outperforms three state-of-the-art baseline methods.

Abstract

It is hard to estimate optical flow given a realworld video sequence with camera shake and other motion blur. In this paper, we first investigate the blur parameterization for video footage using near linear motion elements. we then combine a commercial 3D pose sensor with an RGB camera, in order to film video footage of interest together with the camera motion. We illustrates that this additional camera motion/trajectory channel can be embedded into a hybrid framework by interleaving an iterative blind deconvolution and warping based optical flow scheme. Our method yields improved accuracy within three other state-of-the-art baselines given our proposed ground truth blurry sequences; and several other realworld sequences filmed by our imaging system.