CMOS image sensor fixed pattern noise calibration scheme based on digital filtering method. (June 2022)

• Record Type: Journal Article
• Authors: Huang, Sichong; Lu, Tiejun; Lu, Zhenlin; Rong, Jinye; Zhao, Xuan; Li, Jing
• Is Part Of: Microelectronics journal. Volume 124(2022)
• Publication Date: 2022-06
• Subjects: FPN -- Anisotropic Gaussian filter -- Statistic filter; Microelectronics -- Periodicals; Microélectronique -- Périodiques; Microelectronics; Electronic journals; Journals - contents and abstracts; Periodicals; 621.3805
• DOI: 10.1016/j.mejo.2022.105431
• Languages: English
• ISSNs: 0959-8324

Abstract

Abstract: In order to improve the removal efficiency and effectiveness of the current fixed-pattern noise, a correction method based on the fuzzy matrix and background features is proposed. Firstly, a particular-shaped Gaussian filter kernel is used to blur the original image and to obtain the fuzzy image. Secondly, the fuzzy image and the original image are statistically extracted and compared. Finally, the denoised results are gained after different rules of processing. Experimental results show that FMASF has a better processing performance in the peak signal-to-noise ratio, structural similarity, and human vision-based image quality analysis methods. Therefore, the algorithm is an effective method for removing fixed pattern noise.