Detection method capable of being used for surface microdefect imaging of stamping parts

The invention discloses a detection method capable of being used for surface micro-defect imaging of stamping parts. The detection method comprises the following steps: S1, collecting generalized phase shift dual-wavelength interferograms of a plurality of micro-defect stamping parts and corresponding single-wavelength interferograms of the micro-defect stamping parts; S2, after the collected generalized phase-shift dual-wavelength interferograms of the micro-defect stamping part and the corresponding single-wavelength interferograms of the micro-defect stamping part are cut, an interferogram data set of the micro-defect stamping part is constructed; S3, designing a micro-defect stamping part dual-wavelength interference diagram decoupling convolutional neural network based on deep learning; and S4, using the microdefect stamping part to interfere the pattern data set. The invention provides a detection method capable of being used for stamping part surface microdefect imaging, so that the method can quickly