基于数学形态学和遗传优化的图像去噪

针对图像滤波时损失图像细节这一问题,提出了一种自适应多尺度形态滤波方法,在普通多尺度形态开、闭滤波基础上增加了多尺度top hat变换和bottom hat变换,用于提取并平滑小尺度的图像信息。top hat变换和bottom hat变换的系数对整个滤波器性能起着重要的作用,采用遗传优化的方法对其进行优化。实验结果表明,该方法噪声去除效果好,图像细节保持完整,提高了输出图像的信噪比,增强了滤波器的自适应性和智能性,处理效果明显优于传统滤波方法。

Noise removal from image data based on mathematical morphology and genetic optimization

Considering the details losing in image filtering, an adaptive multiscale morphological filtering (AMMF) method is presented. The proposed method adds the multiscale top-hat transformation and bottom-hat transformation to the conventional multiscale morphological opening and closing filtering. The two added transformations are used to extract and smooth the features which are smaller than the current scale. The coefficients of the multiscale top-hat transformation and bottom-hat transformation have great effects on the performance of the whole filtering. So, they are optimized by aid of the genetic optimization method. Experiment results demonstrate that the AMMF method can remove noise effectively and preserve the details of images completely. It promotes the adaptability and intelligence of the filtering, and manifests better performances than the conventional filtering methods.