摘要
引入了基于提升法的自适应离散小波变换 ,根据LMS自适应法确定伯恩斯坦预测算子的权重系数 ,使其自适应匹配特定的数据序列 ,而且应用该方法结合软域值可实现信号去噪 ,最后扩展该方法应用于二维图象的去噪 ,数值实验表明自适应提升小波变换有效地实现了图象的去噪而且保持了图像的边缘和纹理特性 。
Adaptive discrete wavelet transform based on the lifting scheme was introduced. The weighed coefficient of Bernstein filter predictor was determined to match adaptively with a desired signal by LMS criteria. The algorithm can be applied to signal denoise by the soft threshold of wavelet. Finally this method was extended to denosing of two dimension images. The numerical experiment shows that the method is a powerful method for denoising image. It can keep the character of edge and texture of the image. The advantages of lifting scheme Lie in its flexible design and simple comput.
出处
《红外与毫米波学报》
SCIE
EI
CAS
CSCD
北大核心
2002年第6期447-450,共4页
Journal of Infrared and Millimeter Waves
基金
国家 8 63高科技 (批准号 92 2 1-0 1)资助项目~~
