摘要
论文提出了一种基于改进遗传算法的图像小波阈值去噪方法。从理论上分析了小波阈值去噪的原理,并采用改进遗传算法来求小波变换各子带的最优阈值,计算时无需噪声方差等先验信息;通过综合交叉和随机变异,避免了人为确定交叉率和变异率,从而使算法更加稳健,在提高搜索效率的同时减少陷入局部最优的机会。实验结果表明,与普通的小波阈值去噪方法相比,该方法能较好地改善去噪后图像的视觉效果,提高峰值信噪比。
This paper presents an image denoising method based on wavelet transform and improved genetic algorithm. Thresholds of every wavelet subband can be obtained without requiring the prior knowledge of the noise variance by using the method of improved genetic optimization.Adopting comprehensive crossover and random mutation in order to avoid the difficulty of confirming crossover probability and mutation probability.The robust algorithm enhances searching efficiency and reduces the chance of converging to a local optimum.The experimental results presented in this paper show that compared with the other common wavelet threshold denoising methods,the presented method can improve the visual effect and increase PSNR of the denoised image.
出处
《计算机工程与应用》
CSCD
北大核心
2005年第22期66-69,共4页
Computer Engineering and Applications
基金
光电技术及系统教育部重点实验室访问学者基金
关键词
遗传算法
综合交叉
随机变异
小波阈值去噪
genetic algorithm,comprehensive crossover,random mutation,wavelet threshold denoising