Abstract For the shortcomings of shorter monitoring distance and low-level automation in vision measurement, we propose a deformation monitoring method based on shape matching, edge and region detection of the target. We design a monitoring mark suitable for the method, as well as corresponding algorithms. We choose SLR cameras to carry out experiments, and use continuous shooting smooth method and matching correction technology. We weaken measurement errors caused by environmental factors and reflector and shutter mechanical movement of camera. The visual deformation monitoring software is developed with VS2015 and QT programming platforms and C++ programming language. Long-term monitoring experiments are carried out. Our research is helpful for considering the feasibility, validity and monitoring precision of deformation monitoring method based on visual measurement.