相关双采样技术作为CCD视频信号标准预处理技术之一, 能有效地抑制CCD复位噪声。双采样时钟位置是相关双采样技术中的关键参数。针对自动确定采样时钟位置这一问题, 提出了类对比度评价函数, 通过固定场景改变信号采样时钟位置和复位采样时钟位置, 最大化类对比度评价函数, 自动确定CCD相关双采样时钟位置, 并对类对比度作为评价函数的原理进行了详细的论证。为了验证算法的有效性, 设计了验证平台。实验表明: 该方法能够有效地确定相关双采样时钟的位置, 而且算法确定的位置不稳定度低至3%,对不同亮度具有较强的鲁棒性, 同时该方法相对于传统方法能够有效提高获取图像的信噪比。

Correlated double sampling is an indispensable processing technique for CCD signal conditioning. It can effectively reduce the reset noise. The double sampling clock phases are critical parameters determining correlected double sampling(CDS) technique′s effectiveness. To automatically determine the CDS clock phases, a pseudo-contrast based method was proposed. Through globally searching the maximum of the pseudo-contrast function by changing the signal sampling location and the reset sampling location in a fixed scene imaging set-up, the CDS clock phases could be determined automatically. Principle of the method was explained in detail. A test bench system was designed to verify the proposed methods. Experimental results show the effectiveness of the proposed method with an uncertainty about 3%, the robustness to different light intensities and better SNR performance compared to the conventional method.