自适应小波阈值去噪算法及在图像处理中的应用  

Self-adaptive Wavelet Threshold De-noising Method and Its Application in Image Processing

作　　者：林杰[1] 付梦印[1] 李道平[1] 

机构地区：[1]北京理工大学自动化学院,北京100081

出　　处：《兵工学报》2011年第7期896-900,共5页Acta Armamentarii

摘　　要：针对图像去噪的问题,提出一种基于自适应小波阈值去噪的算法。该方法根据噪声在小波变换下的特性自适应确定小波分解的阈值,并将去噪图像的峰值信噪比作为性能指标,采用中点法进行寻优,得到最优的阈值参数。该方法具有自适应性强、算法简单、去噪效果好等特点。通过实例表明,该方法比原有的阈值去噪算法去噪效果更佳,使图像达到更良好的视觉效果。For the problem image de-noising,a new method based on self-adaptive wavelet threshold algorithm was proposed.It can decide the threshold of wavelet analysis automatically according to the noise characteristics in wavelet transform,then takes PSNR as a function of filter parameter and acquires the optimal threshold parameter by using midpoint method.It has the strong self-adaptation,simple calculation,good de-noising result and so on.Experiment results show that this new method is much better than original threshold de-noising algorithms and achieves better visual effect.

关 键 词：信息处理技术 自适应 小波变换 图像去噪 中点法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]