采用频域融合方法的砂轮刀具磨损三维重构技术
Three？？Dimensional Reconstruction of Tool Wear Area for Grinding Wheel Using Frequency？？Domain Fusion Method

针对聚焦合成方法在砂轮刀具磨损区域三维重构过程中高度值离散的问题,运用傅里叶变换将聚焦合成方法与明暗形状恢复方法相结合,提出了聚焦合成方法的高度值连续化新方法。运用傅里叶变换,在频域中对聚焦合成方法形成的三维重构图像进行高通滤波,对明暗形状恢复方法形成的三维重构图像进行低通滤波,并将滤波后的图像在频域叠加,经过傅里叶反变换,得到空间域融合后的刀具磨损区域三维重构图像。分析和实验表明,融合后的三维重构图像同时保留了聚焦合成方法中蕴含的离散几何高度信息和明暗形状恢复方法中蕴含的表面细节特征信息,能有效避免由于聚焦合成方法中高度值离散带来的误差。
Aiming at the problem of the discontinuous height values in the three？？dimensional reconstruction of tool wear area for grinding wheel using the method of depth from focus (DFF), a new method for continuing the height values with DFF is proposed by using Fourier transform to combine the methods of DFF and shape from shading (SFS). Using Fourier transform in the frequency domain, performing highpass filtering for the 3D reconstruction image of DFF and lowpass filtering for the 3D reconstruction image of SFS, and superimposing the filtered image in the frequency domain, the 3D reconstruction image of the tool wear area of grinding wheel can be obtained by inverse Fourier transform to fuse in the spatial domain. Analysis and experimental results show that the fused 3D reconstruction image not only retains the discrete geometric height information contained in DFF and the detailed surface information contained in SFS, but also effectively avoids the errors caused by the discontinuous height values in the DFF