Fast Approximate Matching of Cell-Phone Videos for Robust Background Subtraction
This addresses a specific instance of background subtraction for user-generated videos, which is incremental as it focuses on handheld camera scenarios rather than a broad new paradigm.
The paper tackles the problem of extracting near-field foreground objects from handheld cell-phone videos by matching two videos of a scene with and without the objects, proposing an efficient spatio-temporal frame matching solution that exploits temporal smoothness. It validates the approach with theoretical error bounds, simulation experiments, and real video tests, showing results compared to alternate methods.
We identify a novel instance of the background subtraction problem that focuses on extracting near-field foreground objects captured using handheld cameras. Given two user-generated videos of a scene, one with and the other without the foreground object(s), our goal is to efficiently generate an output video with only the foreground object(s) present in it. We cast this challenge as a spatio-temporal frame matching problem, and propose an efficient solution for it that exploits the temporal smoothness of the video sequences. We present theoretical analyses for the error bounds of our approach, and validate our findings using a detailed set of simulation experiments. Finally, we present the results of our approach tested on multiple real videos captured using handheld cameras, and compare them to several alternate foreground extraction approaches.