Towards an Automated Image De-fencing Algorithm Using Sparsity

TL;DR

This paper introduces an automated image de-fencing method for dynamic scenes, combining fence detection, motion estimation, and multi-frame data fusion within an optimization framework to improve robustness over traditional approaches.

Contribution

It presents a novel automatic de-fencing algorithm that handles dynamic scenes by integrating fence detection, motion estimation, and data fusion using an optimization approach.

Findings

Effective fence detection using Gabor filter and machine learning.

Successful removal of fences in dynamic scenes.

Robust de-fencing demonstrated through experimental results.

Abstract

Conventional approaches to image de-fencing suffer from non-robust fence detection and are limited to processing images of static scenes. In this position paper, we propose an automatic de-fencing algorithm for images of dynamic scenes. We divide the problem of image de-fencing into the tasks of automated fence detection, motion estimation and fusion of data from multiple frames of a captured video of the dynamic scene. Fences are detected automatically using two approaches, namely, employing Gabor filter and a machine learning method. We cast the fence removal problem in an optimization framework, by modeling the formation of the degraded observations. The inverse problem is solved using split Bregman technique assuming total variation of the de-fenced image as the regularization constraint.