多尺度小波降噪的数字散斑相关搜索

Digital speckle correlation method of multi-scale wavelet noise reduction

提出了一种基于多尺度小波降噪的数字散斑相关搜索方法。选用symlets小波,对分别存在高斯白噪声、椒盐噪声及泊松噪声的散斑位移图像进行多级小波分解,采用不同的降噪策略处理后再进行相关搜索。计算结果表明,多尺度小波降噪的数字散斑相关搜索方法与传统空域相关搜索方法相比,其测量精度提高了一个数量级,相对误差可以控制在1%以内;同时,其计算效率提高了1倍。

A novel Digital Speckle Correlation Method （DSCM） based on multi-scale wavelet noise reduction is proposed. Speckle patterns with Gaussian white noise or Salt ＆ pepper noise or Poisson noise are decomposed using symlets wavelet family and processed by different noise reduction strate gies. Compared with the traditional DSCM, the accuracy of this new method is improved dramatically and the relative error is less than 1%. Meanwhile, the calculated consuming time is decreased to half of the traditional DSCM.