高速CCD的CDS控制参数自适应算法研究

Adaptive Optimization Agorithm for CDS Control Parameters of High-speed CCD

优化相关双采样(CDS)控制参数可以提高电荷耦合器件(CCD)输出信噪比,但是高速CCD的CDS控制参数很难通过实验确定。为了得到较好的CCD输出信噪比,提出了一种CDS控制参数自适应算法:在拍摄目标和曝光参数不变的情况下,改变复位脉冲宽度、噪声采样时刻、数据采样时刻3个CDS控制参数,对每种参数组合采集一帧图像,计算各帧图像中每个像素的Tenengrad函数,将每帧像素的Tenengrad平方和作为清晰度评价参数,选择清晰度评价参数最大值对应的复位脉冲宽度、噪声采样时刻、数据采样时刻作为CDS控制参数,就可使CCD输出达到采用CDS技术下最好信噪比。根据所提出的方法,在试验电路中,由现场可编程门阵列(FPGA)实现了自适应算法,并进行了清晰度对比测试。测试表明,与没有采用自适应算法的最好图像相比,测试参数Edge Profile从单位影像高度507.4像素提高到763.8像素,测试参数MTF 50从526.5 LW/PH提高到937.8 LW/PH。试验结果表明,应用自适应算法可以有效提高CCD的输出信噪比。

The signal-to-noise ratio of Charge Coupled Devices（CCD） output can be improved by the optimization of control parameters of Correlated Double Sampling（CDS）.However,the control parameters of CDS of high-speed CCDare difficult to determine from experiments.An adaptive algorithm was proposed to get a better signal to noise ratio of CCD output.With fixed target and exposure parameters,the CDS control parameters such as the width of reset pulse,the timing of noise sampling and data sampling were tuned.The images on all combination of the CDS control parameters were acquired.The square sum of the Tenengrad function of all pixels on each image was used to evaluate the quality of the image.The combination of three CDS parameters corresponding to the maximum sum was chosen as CDS control parameters,resulting the best signal-to-noise ratio of the CCD out.Finally,the algorithm was implemented in Field-Programmable Gate Array（FPGA） and the experiments results showed that compared with the best image acquired without adaptive algorithm,the parameters of Edge Profile are improved from 507.4 pixels per PH to 763.8 pixels per PH,the parameters of MTF50 improved from 526.5 LW/PH to 937.8 LW/PH.It showed that adaptive algorithm can effectively improve the signal to noise ratio of CCD output.