基于遗传算法的多尺度小波阈值去噪方法

Multi-scale wavelet threshold denoising method based on genetic algorithm

针对传统小波硬、软阈值去噪方法分别存在视觉失真、边缘模糊等缺点,采用一种折衷的阈值去噪方法,并结合图像本身的特征和多尺度小波变换的优点,提出一种基于遗传算法的多尺度小波阈值去噪方法,并且为提高遗传算法的性能,在传统遗传算法的变异操作后增加种群更新机制.结果表明,该方法可以有效地去除高斯白噪声干扰,较好地保留图像的边缘信息,去噪效果较好.

In view of the fact that tradition wavelet hard and soft threshold denoising method suffers from the defect of visual distortion and edge blurring,we put forward an eclectic threshold denoising method,combined with the characteristics of image and multi-scale wavelet transform of advantages,based on the genetic algorithm multi-scale wavelet threshold denoising method to improve the performance of genetic algorithm,the traditional genetic algorithm in the variation operation after increase update mechanisms. Experimental results show that this method can effectively remove the Gaussian white noise interference and good image edge information retained,achieving a good effect of denoising.