Multi-parameter Tikhonov regularization method based on machine learning

The invention provides a multi-parameter Tikhonov regularization method based on machine learning, and the method is characterized in that the method comprises the following steps: S1, data collection: determining a research object, and collecting a picture meeting a scene; S2, preprocessing the image; S3, calculating an optimal Tikhonov regularization parameter; S4, performing image restoration;S5, analyzing a result according to the Tikhonov regularization parameter and the relative error rate. According to the method, targeted optimization is carried out on the images meeting the boundaryconditions of the same kind of period, and the optimal Tikhonov regularization parameter vector conforming to the images of the kind is calculated in advance through preprocessing and an optimized andimproved machine learning algorithm. When the damaged image is input, the image can be recovered quickly and efficiently, and compared with a traditional scheme, the method improves the image recovery quality. 本发明提供1、一种基于机器学习的