Leaf area index (LAI) is of great significance for crop growth monitoring, agricultural disaster stress monitoring and yield prediction. There are three red-edge bands (705nm, 740nm, 783nm) for Sentinel-2 satellite images, which are very sensitive to vegetation growth. Unfortunately, the spatial resolution of these three red-edge bands (20m) is inconsistent with that of visible and near infrared bands (10m), which limits the application of Sentinel-2 images. For solving this problem, the six bands with spatial resolution of 20m was reconstructed into the spatial resolution of 10m by using super-resolution for multispectral multiresolution estimation (SupReMe) algorithm. Using the reconstructed Sentinel-2 image, the corn canopy LAI was retrieved by using the PROSAIL radiative transfer model and the random forest machine learning method. The results showed that the space details of Sentinel-2 image were improved while the spectral invariance was maintained after reconstruction by using SupReMe algorithm. The determination coefficients (R2) of LAI retrieving using reconstructed image was improved from 0.70 to 0.68 compared with resampling Sentinel-2 image, and the root mean square error (RSME) was improved from 0.240 to 0.262. The results showed that the SupReMe method can be used to reconstruct the spatial resolution of Sentinel-2 image and the reconstructed image can be used to improve corn canopy LAI retrieving accuracy.