基于行过扫数据降低大面阵CCD图像噪声算法研究

Algorithms to Reduce Noise in CCD Using Line Overscan Data

摘要:

针对全帧转移大面阵电荷耦合器件(CCD)多帧图像数据随时间漂移的现象和测试电路引入的噪声,导致无法准确评估CCD器件的参数,文章提出一种基于行过扫数据校正算法。行有效像元数据与对应行垂直过扫数据同时输出,通过有效像元数据减去垂直过扫数据均值,对图像数据漂移的现象进行抑制。测试电路中模拟模块与数字模块同时作用于单行有效像元数据与垂直过扫数据,通过有效像元数据减去垂直过扫数据均值,消除测试电路引入的噪声。实验结果表明,经过该算法校正后的器件读出噪声减小了20%,暗场差值图像中超过15e-和25e-的比例减小了25%。该算法适用于大面阵CCD器件,可校正测试系统引入的测试误差,提高全帧转移大面阵CCD器件测试效率。

Abstract:

This paper presents a correction algorithm based on line overscan data to solve the testing error produced by drifting image data in a full-frame transfer large-array charge-coupled device (CCD) and the noise introduced by the testing circuit, which prevent the accurate evaluation of the parameters of a CCD device. The effective image data and vertical overscan data were output simultaneously. The drifting image data were suppressed by subtracting the mean value of the vertical overscan data from the image data. In the testing circuit, the analog and digital modules simultaneously acted on a single line of the image data and vertical overscan data. The noise introduced by the testing circuit was eliminated by subtracting the mean value of the vertical overscan data from the image data. Experimental results showed that the proposed algorithm reduced the readout noise of the device by 20%. The ratio for values greater than 15e- and 25e- in a dark difference image was reduced by 25%. The algorithm is suitable for a large-area-array CCD device and can correct the errors introduced by the testing circuit to improve the testing efficiency of a full-frame transfer large-area-array CCD device.