GENERALIZED POLYHEDRAL OBJECT RECOGNITION AND LOCALIZATION USING CROSSBEAM SENSING

Object localization is a fundamental task in industrial automation. In this article, we present a recognition and localization technique bas ed upon binary beam sensors. Binary beam sensors, which consist of mod ulated LED light sources and detectors, are appropriate for manufactur ing applications due to their reliability, high accuracy, robustness, inexpensiveness, and ease of calibration. By extracting sufficient inf ormation, recognition and localization can be performed as fast (appro ximate to 0.1 s) and as accurately (approximate to 25 microns) as high -speed manufacturing requires. Generalized polyhedral objects are reco gnized and localized by being passed through a crossbeam sensor, which is a set of coplanar oriented binary light-beam sensors; robot positi ons are recorded when the beam sensors' outputs change. Recognition an d localization share two subproblems: the correspondence problem (the problem of interpreting the sensed features in terms of the model feat ures), and the pose-estimation problem (the problem of estimating the object's pose from the sensed data and an interpretation of the sensed features in terms of model features). In Wallack and Manocha (1994), we described a pose-estimation technique for crossbeam sensor data. In this article, we present two methods for solving the correspondence p roblem for crossbeam sensor data: a linear time completely on-line met hod, and a constant-time on-line method that utilizes preprocessing.