Research On Steel Plate Appearance Quality Detection Method Based On Grating Projection

The appearance quality inspection of steel plate can be divided into the surface defect detection of steel plate and the thickness detection of steel plate.At present,the mainstream steel plate defect detection method can not detect the thickness of steel plate and the surface defect of steel plate at the same time.Aiming at this problem,this paper uses Raster Projection 3D profilometry to apply it to the inspection of steel sheet appearance quality,so as to achieve the purpose of detecting the thickness of steel plate and surface defects in the same time.The work of this article is as follows:This paper improves the triangular imaging method of grating projection three dimensional profilometry,making the distortion of the moire image captured by the CCD camera small.On this basis,the whole system is calibrated to eliminate the tangential distortion and radial distortion caused by CCD camera lens.Finally,the moire image captured by the CCD camera is corrected to eliminate the distortion of the gray level curve caused by ambient light.In order to get the best effect of phase extraction,wavelet transform profilometry based on Morlet wavelet is used to get the phase value of moire image.By analyzing the advantages and disadvantages of row by point expansion algorithm,branch cutting method and mass map guidance method,the branch cutting method is chosen as the algorithm of truncated phase expansion.Finally,on the basis of the Gauss low pass filter,the surface of the reconstructed steel plate is filtered,which can effectively reduce the surface error of the steel plate caused by the reconstruction algorithm.Experiments show that the accuracy of steel plate thickness measurement is 0.03 mm,and the surface defects of steel plates with minimum size 0.3mm can be detected.