The position and shape information of defects can be determined by using ultrasonic guided wave to invert the defects in the plate structures. Aiming at the problem that the travel time imaging method is not effective in the low frequency range， an ultrasonic guided wave defect inversion imaging method based on wave field is proposed. According to the wave equation， the imaging principle of beamforming is derived， and the scattering field data in Born approximation are used to coherently stack the values of pixels in the imaging area to obtain the position and shape of defects. Through the mapping relationship between beamforming and Fourier diffraction theorem in frequency domain， the results of beamforming are transformed into diffraction tomography images， and some clearer inversion images are obtained. Aiming at the problems of artifact and noise in diffraction tomography， the two-dimensional variational mode decomposition （2D-VMD） method is used to denoise the images， which can effectively remove the artifact and the burr of defect contour edge， and further improve the imaging resolution. The inversion results show that the proposed method can accurately reconstruct the location， size and shape of the thinning defects on the aluminum plate with high resolution.