Automatic moving foreground extraction using random walks

Abstract

In this paper, we propose a novel approach for automatic foreground extraction in video frames by analyzing the spatiotemporal aspect. We divide our contribution to tree steps: Automatic seeds detection, formulating the energy function, and using the random walk algorithm to minimize this function. First, we detect seeds by extracting a sparse of good features to track in the current frame and compute the difference between those pixels and its adjacent in the previous frame, the difference of pixels is treated in HSV color space to make the result more accurate, we thresholds this difference, and we classify moving and stationary pixels. Secondly, we formulate our foreground extraction as a graph based problem, then we define an energy function to evaluate spatiotemporal smoothness. Finally, we applied the random walk algorithm with seeds detected in the first step to minimize the energy function problem, the solution leads to evaluate the potential that every pixel in the video sequences is marked in motion or a stationary pixel. We suggest that our unsupervised method has the potential to be used for many kinds of motion detection and real-time video.