GEOMETRIC, ALGEBRAIC, AND THERMOPHYSICAL TECHNIQUES FOR OBJECT RECOGNITION IN IR IMAGERY

We describe a new approach for computing invariant features in infrare d (IR) images, Our approach is unique in the field since it considers not just surface reflection and surface geometry in the specification of invariant features, but it also takes into account internal object composition and thermal state which affect images sensed in the nonvis ible spectrum. We first establish a nonlinear energy balance equation using the principle of conservation of energy at the surface of the im aged object, We then derive features that depend only on material para meters of the object and the sensed radiosity. These features are inde pendent of the scene conditions and the scene-to-scene transformation of the ''driving conditions'' such as ambient temperature and wind spe ed. The algorithm for deriving the invariant features is based on the algebraic elimination of the transformation from the nonlinear object- image relationships. The elimination approach to compute absolute inva riants is a general method based on deriving and separating the polyno mial relations which are invariant with respect to the given transform ation. A complete model-based approach for recognition of objects in I R images is presented. Geometric invariant features are used to genera te hypotheses of object identity and pose. These hypotheses are verifi ed or refuted by the thermophysical features. Results on real IR image ry are shown to illustrate the performance of the features and the met hodology in an object recognition system that deals with multiple clas ses of objects. (C) 1998 Academic Press.