基于多级维纳滤波的OCT图像除噪方法研究

为了实现光学相干层析成像（OCT）过程中采集信号的噪声去除，设计了基于维纳滤波和对比度增强的OCT图像处理方法。首先对OCT图像的噪声进行了分析，继而针对OCT信号的噪声干扰问题提出一种除噪算法，通过多尺度维纳滤波器对图像进行滤波处理，然后根据区域特性采用对比度增强方法提高图像对比度。实验结果表明：经该方法处理后图像的背景方差（BV）较原始图像降低了3-7倍，而且信噪比高，细节方差和背景方差之比（DV/BV）提高了3-6倍。该方法不仅有效去除了OCT图像的噪声，而且视觉效果良好，是一种有效的图像后处理方法。

An OCT Image Denoising Method Based on Multi-scale Wiener Filtering

In order to realize OCT (Optical Coherence Tomography) image denoising, an image denoising method is proposed and its applied algorithms such as Wiener filtering and contrast enhancement. First, analyzed the OCT image noise, then proposed an multi-scale Wiener filtering algorithm based on character of noise in the image. The method denoises OCT image gradually by different Wiener filter template. Finally, the contrast enhancement based on the character of the different areas is used to strengthen the image information. Experiment results demonstrate that the algorithm enhances OCT image the value of the signal to noise ratio with the background variance (BV) of the processed image was higher 3-7 times than the original image, and the value of DV/BV was higher 3-6 times, It can get the conclusion that the method denoised OCT image with good visual effect, and it is a effective image processing method.