中文摘要：

暂态噪声是CMOS图像传感器暗光下噪声的重要组成部分,不但影响图像质量,而且限制图像动态范围的提高。提出了一种新的自适应片上抑制CMOS图像传感器暂态噪声的方法。将多帧采样图像进行空间转换,使用自适应增益将多帧采样图像中相应像素进行加权平均,进行空间逆变换,得到去噪后的像素值。对20组图片进行实验,结果表明算法能够克服传统基于多帧采样暂态噪声去除方法存在的改变原图像亮度的缺点,所得PSNR值和占用逻辑单元数量均优于其它方法,最高频率可达140MHz。并且分析了采样图像帧数对去噪效果的影响,得到最优采样帧数,为消除CMOS图像传感器暂态噪声提供一种新的解决思路。

英文摘要：

Temporal noise is the important component of the dark noise of CMOS Image Sensor.It not only will deteriorate the image quality,but also restrain improving the dynamic range of the image.In this paper,an adaptive on-chip denoising method for temporal noise of CMOS Image Sensor is proposed.Spatial transform is employed to the multi-sample images,the corresponding pixels of multiple sample images are weighted average by the adaptive gain.Then spatial inverse transformation is adopted to obtain the denoised pixel values.20 groups image have been tested,the experimental results show that the proposed method can overcome the drawback of traditional method based on multi-sample,come up to higher PSNR value and take up less logical units than other methods,the max frequency is about 140MHz.This paper also analyzes the effect of sampling frames to the denoising results,and obtain the optimal sampling frame.The proposed method supplies a new solution for eliminating the temporal noise of CMOS Image Sensor.