Chromatic aberration correction in RAW domain for image quality enhancement in image sensor processors

Abstract

Taking sharp and high-resolution images usually relies on high quality camera systems. In this paper, we present an approach to correct chromatic aberrations distortions on images caused by imperfect optics. The proposed system relies on a calibration step where the distortion produced by the lens is estimated, followed by a second stage where the image is enhanced. The proposed system works in RAW domain as a pre-processing process to facilitate the demosaicing, and noise reduction process in an image sensor processor (ISP). The results of the proposed method prove that both the lateral and the longitudinal aberrations can be corrected, is computational efficient and also facilitates the functioning of an ISP pipeline producing sharper and less noisy images.