A novel technique for reducing demosaicing artifacts

D. Menon, S. Andriani, and G. Calvagno

Proc. of the 14th European Signal Processing Conference (Eusipco), Sept. 2006

[Articlepdf]   [BibTeX]  
Abstract: Demosaicing is the process of reconstructing the full dimension representation of an image captured by a digital camera with a color filter array. The color filter array allows for only one color measuring for each pixel and the missing two colors have to be estimated. In literature many demosaicing techniques have been proposed but the reconstructed images are affected by some visible and annoying artifacts. In this paper we propose a new effective algorithm to reduce these artifacts. This algorithm improves the performances of the demosaicing reconstruction, increasing the visual quality of the resulting images. It can be applied directly after the color interpolation, or as a off-line post-processing to improve the image acquired by the digital camera.

In this site various demosaicked images refined with the algorithm proposed in the paper are available for an analysis of its performance.

The original images are taken from www.cipr.rpi.edu. We downsampled them according to the Bayer pattern and we reconstructed them with some known demosaicing techniques. Afterwards we applied the proposed algorithm to reduce the demosaicing artifacts.

Go to a visual comparison of some resulting images (the PNG graphic format is used).

Click here to download the demosaicked images (the uncompressed TIFF format is used).

We compared the performance of the refining algorithm over the following demosaicing approaches:

Instead, if you search a new demosaicing algorithm, you can see Demosaicing with directional filtering and a posteriori decision.