A vision-based algorithm for estimating tip interaction forces on a deflected Atomic Force Microscope (AFM) cantilever is described. Specifically, we propose that the algorithm can estimate forces acting on an Atomic Force Microscope (AFM) cantilever being used as a nanomanipulator inside a Scanning Electron Microscope (SEM). The vision based force sensor can provide force feedback in real-time, a feature absent in many SEMs. A methodology based on cantilever slope detection is used to estimate the forces acting on the cantilever tip.