摘要
为了获得更清晰的X射线图像,在去噪的同时尽可能保护边界完整,针对传统的均值滤波去噪时造成边缘模糊的问题,提出了一种改进的邻域平均滤波去噪方法。对图像中的未知噪声进行检测分析,确定噪声类型和参数;使用改进的邻域平均滤波方法对图像去噪,对滤波模板邻域内像素与中心像素做差值,针对不同类型噪声,选择不同数量的邻域内对应差值较小的几个像素取平均值替代中心像素,避免不同区域像素被混叠处理造成边界模糊;计算去噪后X射线图像的均方误差和峰值信噪比,对去噪效果进行客观评价,并将提出的去噪方法与常见的滤波方法进行比较。实验结果表明,改进的邻域平均滤波去噪方法相比常见的滤波方法在去噪的同时能够更好地保护边界。
In order to obtain a clearer X-ray image and protect the boundary integrity as much as possible while denoising, an improved denoising method by neighborhood mean filtering is proposed to solve the problem of blurred edges caused by traditional mean filtering. We detect and analyze the unknown noise in the image, determine the noise type and parameters;use the improved neighborhood average filtering method to denoise the image, and make the difference between the pixels in the neighborhood of the filter template and the center pixel, and choose different values for different types of noise. The average value of several pixels with small corresponding differences in a number of neighborhoods is used to replace the center pixel, so as to avoid the blurring of boundaries caused by the aliasing of pixels in different regions;the mean square error and peak signal-to-noise ratio of the denoised X-ray images are calculated, the denoising effect is objectively evaluated, and the proposed denoising method is compared with the common filtering methods. The experimental results show that the improved neighborhood average filtering denoising method can better protect the boundary while denoising than the common filtering method.
出处
《图像与信号处理》
2023年第1期1-8,共8页
Journal of Image and Signal Processing
