Color Demosaicing Using Variance of Color Differences

Abstract

Abstract-This paper presents an adaptive demosaicing algorithm. Missing green samples are first estimated based on the variances of the color differences along different edge directions. The missing red and blue components are then estimated based on the interpolated green plane. This algorithm can effectively preserve the details in texture regions and, at the same time, it can significantly reduce the color artifacts. As compared with the latest demosaicing algorithms, the proposed algorithm produces the best average demosaicing performance both objectively and subjectively.

Reduction of Colour Artifacts Using Inverse Demosaicking

“…They were then demosaicked using two chosen simple demosaicking methods, namely Bilinear and Freeman [3]. After the application of our proposed re-demosaicking method, the results are tabulated in Table 1, which gives the Peak Signal-toNoise Ratio (PSNR) [2] and the Normalized Color Difference (NCD) [12] quality measures. This confirms that our proposed method produced a re-demosaicked image with a quality higher than its input.…”

“…In order to compare the energy levels of the high frequency components between the two sub-images, we use filter banks to perform an undecimated wavelet transform. The filter bank structure consists of 2D seperable filters constructed from a low-pass filter (1) and a high-pass filter (2) to decompose each sub-images into 4 subbands, namely (LL) both rows and columns are low-pass, filtered, (LH) rows are low-pass filtered, columns are high-pass filtered, (HL) rows are high-pass filtered, columns are low-pass filtered, (HH) both rows and columns are high-pass filtered.…”

“…With the advancement in the development of demosaicking algorithms [4,9,5,8,2,13,7], high quality images with minimal artifacts can be produced. In this paper, we propose to take advantage of this new development to re-demosaicked older images with colour artifacts which were previously produced by inferior primitive demosaicking techniques.…”

“…Our proposed method is to extract those original sensor data from the full colour image for re-demosaicking using state-of-the-art demosaicking algorithms [4,9,5,8,2,13,7]. In section 2, we propose an inverse demosaicking technique to first extract the original CFA sensor data from the stored full colour image based on the assumption that the CFA sensor data remains more-or-less intact in the image processed by the simple demosaicking technique implemented in the digital camera.…”

An Efficient Combined Demosaicing and Zooming Algorithm for Digital Camera

“…1. To convert them to full-color images, the two missing color components of each pixel have to be estimated by color demosaicing [2][3][4][5][6][7].…”