基于Curvelet变换的ESPI图像散斑噪声抑制算法研究

摘要

This paper studies the de-noising principle of curvelet transform. Combined with the characteristics of ESPI fringe pattern,the improved Curvelet de-noising method is proposed. That is,making the logarithmic transformation to the image at first,then the multiplicative noise is changed to additive noise. After that,the Curvelet transform is made, which is beneficial to separate the noise and edge information. The experimental results show that this method is effective to remove speckle noise,and simultaneously keep fringe pattern edge information.%重点研究了Curvelet变换的去噪原理及方法,并结合ESPI条纹图中散斑乘性噪声的特性,提出了改进的Curvelet变换去噪方法,即在去噪之前,先将图像进行对数变换,使得乘性噪声变成加性噪声,然后再进行Curvelet变换去噪,这样会更有利于噪声与边缘信息的分离.并把该方法应用在对轮胎的激光无损检测图像处理系统中,实验结果表明该方法在有效去除散斑噪声的同时,较好地保持了条纹图的边缘细节信息.