PROJECT TITLE :

Removing Camera Shake via Weighted Fourier Burst Accumulation - 2015

ABSTRACT:

Varied recent approaches attempt to get rid of image blur because of camera shake, either with one or multiple input pictures, by explicitly solving an inverse and inherently unwell-posed deconvolution problem. If the photographer takes a burst of images, a modality obtainable in just about all trendy digital cameras, we tend to show that it's doable to mix them to induce a clean sharp version. This is completed without explicitly solving any blur estimation and subsequent inverse drawback. The proposed algorithm is strikingly simple: it performs a weighted average in the Fourier domain, with weights relying on the Fourier spectrum magnitude. The method can be seen as a generalization of the align and average procedure, with a weighted average, motivated by hand-shake physiology and theoretically supported, happening within the Fourier domain. The strategy's rationale is that camera shake has a random nature, and so, each image within the burst is generally blurred differently. Experiments with real camera data, and extensive comparisons, show that the proposed Fourier burst accumulation algorithm achieves state-of-the-art results an order of magnitude faster, with simplicity for on-board implementation on camera phones. Finally, we also gift experiments in real high dynamic range (HDR) scenes, showing how the method will be straightforwardly extended to HDR photography.