This paper describes an object recognition system for use in complex imagery that can perform recognition adaptively by setting the matching threshold such that the probability of a false positive is low. In order to accurately model small, irregularly shaped objects, we represent the objects using dense sets of edge pixels with associated local orientations. Three-dimensional objects are modeled by a set of two-dimensional views of the object. We allow translation, rotation, and scaling of the views to approximate full three-dimensional motion of the object. We use a version of the Hausdorff measure to determine which positions of an object model are good matches to an image. These positions are determined efficiently through the examination of a hierarchical cell decomposition of the transformation space, which allows large volumes of the space to be pruned quickly. Additional techniques are used to decrease the computation time of the method when matching is performed against a catalog of object models. We then describe a new model of the matching process that allows the probability of a false positive to be estimated efficiently at run-time. Finally, we give results of this system recognizing object in infrared and intensity images.