Title :
Blur-free high-sensitivity imaging system utilizing combined long/short exposure green pixels
Author :
Sato, Satoshi ; Okada, Yusuke ; Azuma, Takeo
Author_Institution :
Adv. Technol. Res. Lab., Panasonic Corp., Kyoto, Japan
Abstract :
In previous work, we have proposed a high-sensitivity imaging system based on the long exposure for all the G(green) pixels, in which the inherent blur was suppressed with using both the regularization method and motion information from blur-free short-term exposed information from R(red) / B(blue) pixels. However this method was ineffective if detected motion information is inaccurate. As a result, an inadequate accuracy of motion information lost image quality and caused blur image. In this study, we propose a new high-sensitivity imaging system that is stable under such a condition. We arrange long- and short-term exposed G pixels for each line alternatively. With combining regularization method and blur-free G pixel information obtained from the short-term exposure, the proposed system does not require motion information. Then reconstruction images of our proposed system include no blur even if accurate motion detection is difficult.
Keywords :
image colour analysis; image restoration; motion estimation; blur image; blur-free high-sensitivity imaging system; blur-free short-term exposed information; combined long/short exposure green pixels; image quality; image reconstruction; motion detection; motion information; regularization method; Computer vision; Image color analysis; Image reconstruction; Imaging; Motion detection; PSNR; Video sequences;
Conference_Titel :
Consumer Electronics (ICCE), 2012 IEEE International Conference on
Conference_Location :
Las Vegas, NV
Print_ISBN :
978-1-4577-0230-3
DOI :
10.1109/ICCE.2012.6161744