Abstract. We address the validation of the current sensed environmental state against a model known from earlier perception. Our surface based approach compares 3D range data of a built environment against an a priori scene model with the analysis of identified differences facilitating the hypothesis of causal structural changes within the scene. Experimental results show good success rates in identifying and analysing realistic structural changes introduced across example industrially themed scenes.
Toby P. Breckon, Robert B. Fisher