It is becoming more and more popular in the cultural heritage field to disseminate cultural relic information using digital media. One of the important aims of digitizing cultural heritage data is to better use the archived digital data to enable the data to fulfill their roles. Using image technology based on analysis, we have proposed a new automated virtual restoration protocol for the Ming Dynasty murals of a Guanyin Temple in Xinjin. We found that the Criminisi algorithm, which is based on the exemplar image restoring model and which uses wavelet energy factors to guide the filling order, can restore information about missing parts very well. It adjusts the restoration rules using appropriate factors for images with different textures. The filling texture patches are from the remaining parts of the same mural, just as with real restoration methods.