像素非均匀性是CCD成像性能评价中的一个重要指标, 反映的是像素结构本身的差异。传统的以灰度值为基础计算DSNU和PRNU的方法未能考虑读出电路引入的时域噪声, 且计算时未剔除不同像素曝光时间不同带来的误差, 计算结果也只适用于某个具体曝光量。在分析CCD信号流的基础上, 厘清灰度值不均匀的影响因素。参考DSNU和PRNU的计算方法, 再结合帧转移型CCD的工作方式, 提出了设置多档曝光时间, 每档曝光时间下采集多帧暗(亮)图像, 再通过拟合求得暗电流(“暗电流+光电流”)后并以之为基础计算暗电流非均匀性DCNU和光电流非均匀性PCNU的方法。同时, 建立CCD像素非均匀性测试系统, 验证其光源的稳定性和均匀性, 以排除采集图像过程中测试系统引入的时域和空间误差。在此测试系统的基础上, 利用新方法测得CCD的DCNU为25.51 e-/pixel·s-1, PCNU为0.98%。相比于传统的DSNU和PRNU值更准确地反映了CCD像素结构的非均匀性, 所得结果更具普遍适用性。

Pixel non-uniformity is an important index in the evaluation of CCD imaging performance, which reflects the difference of pixel structure itself. The traditional method of calculating DSNU and PRNU based on gray value fails to consider the temporal noise introduced by the readout circuit, and the error caused by different pixel′s different exposure time isn′t excluded, the calculation results were also only applicable to a specific exposure condition. Based on the analysis of CCD signal flow, the influence factors of gray value non-uniformity were clarified. Referring to the calculation method of DSNU and PRNU, combined with the working mode of frame-transfer CCD, a new method which calculated the dark current non-uniformity DCNU and the photo current non-uniformity PCNU was put forward. The method including setting multiple exposure time, and the multi-frame dark (bright) images was collected under each exposure time, then fitting for the dark current ("dark current+photo current"). Meanwhile, a CCD pixel non-uniformity test system was established, the stability and uniformity of the light source were verified to exclude the temporal and spatial error introduced by the test system when collecting images. On the basis of the test system, the DCNU 25.51(e-/pixel·s-1), PCNU 0.98% of CCD were measured using the new method. The result more accurately reflects the non-uniformity of CCD pixel structure, and was more universally applicable, compared to traditional DSNU and PRNU value.