一种无阈值的傅里叶窗口滤波算法

提出了一种无需设定阈值的傅里叶窗口滤波算法，该方法解决了加窗傅里叶变换法无法克服的阈值问题，得到单一最佳频率图像，并大大缩短了计算时间．该方法可用于各类条纹图像的低通滤波，为从条纹图中进一步提取变形相位奠定了良好的基础．笔者介绍了自动窗口傅里叶变换算法的原理，利用该算法对散斑图进行了处理，并与加窗傅里叶变换法的滤波结果进行了对比．实验结果证明了无需设定阈值的傅里叶窗口滤波算法具有良好的滤波效果．

A NEW AUTO WINDOWED FOURIER MATCHING METHOD WITHOUT THRESHOLD

A new windowed Fourier transform algorism that does not need to set a threshold in advance called Auto Windowed Fourier Matching algorism is introduced. The only optimum frequency obtained in the local area is the aim of the new algorism, and the computing time is short. The new algorism can be applied to many types of low pass filtering process. This algorism provides a good basis for further extracting deformation phase from speckle fringe patterns. The principle of auto windowed Fourier matching algorism is also introduced, and it is applied to filtering the same fringe pattern in order to compare with the windowed Fourier transform algorism. The experiment results show that the new auto windowed Fourier matching method has a better performance in fringe pattern filtering.