小波消噪在三维测量轮廓术中的应用

Waveletde-noiseing application in3-D transform profilometry

在测量轮廓术中 ,噪声的影响会给测量得到的光强数据带来误差 ,通过带有误差的光强数据恢复出来的物体轮廓数据也有误差 ,而且 ,较高的噪声会让轮廓术中的相位解包裹算法失效 .文中提出先利用小波进行前期处理 ,消除去噪声 ,再进行恢复其轮廓 .计算机模拟结果表明 ,和传统的频域滤波相比 ,应用小波消噪的方法可以使相位误差大为减小 。

In 3- D transform profilometry,there is always white gauss noise,this kind of noise will lead to the measurementerrors ofthe lightintensity and the phase evaluation errors,and also itmakes unwraping algorithm disable.By using the routine filter,it can reduce the noise.In this paper,we apply wavelet method to eliminate the noise firstly.Compared with ordinary filtering method,wavelet de- noising based on Donoho method can eliminate the noise better and build up the robustness of unwarping algorithm.At last,a simulation based on Fourier transform profilometry shows that the wavelet de- noising method can reduce noise more litter and buildup the algorithm more robuster than the routing filter.