To　measure　the　velocity　for　the　omni-directional　intelligent　wheelchair　for　the　depth　research　on　control　and　improve　the　measured　accuracy　and　computational　efficiency,　the　variational　optical　flow　based　velocity　estimation　method　is　used　in　this　paper.　Firstly,　an　improved　energy　functional　model　which　combines　a　brightness　constancy　assumption,　a　gradient　constancy　assumption,　a　Laplacian　constancy　assumption　and　a　discontinuity-preserving　spatio-temporal　smoothness　constraint　is　employed　to　improve　the　accuracy　of　optical　flow　estimation;　Then,　a　planar　optical　flow　model　based　improved　voting　Random　Sample　Consensus(RANSAC)　method　is　presented　to　remove　heterogeneous　optical　flow　vectors　produced　by　non-uniform　ambient　illumination,　local　motion　blur　and　chaotic　surrounding　environment.　Finally,　aiming　at　the　installation　attitude　of　the　camera,　an　effective　method　of　camera　attitude　adjustment　is　introduced　to　simplify　the　spatial　coordinate　system　transformation　and　solve　the　model.　This　paper　has　realized　the　velocity　estimation　of　the　wheelchair　by　real-time　acquisition　of　images　in　the　framework　of　the　Robot　Operating　System(ROS).　Experimental　results　showed　that　the　wheelchair　velocity　estimation　method　performed　robustly　in　the　presence　of　chaotic　surrounding　environment　through　the　above　improvements.