Fusing range and intensity images for mobile robot localization

Abstract

Abstract-In this paper, we present the two-dimensional (2-D) version of the symmetries and perturbation model (SPmodel), a probabilistic representation model and an EKF integration mechanism for uncertain geometric information that is suitable for sensor fusion and integration in multisensor systems. We apply the SPmodel to the problem of location estimation in indoor mobile robotics, experimenting with the mobile robot MAC-ROBE. We have chosen two types of complementary sensory information: 1) range images; 2) intensity images; obtained from a laser sensor. Results of these experiments show that fusing simple and computationally inexpensive sensory information can allow a mobile robot to precisely locate itself. They also demonstrate the generality of the proposed fusion and integration mechanism.Index Terms-Computer vision, map-based localization, mobile robots, range finder, sensor fusion.

Sensordatenfusion zur Robusten und Präzisen EKF Lokalisierung von Mobilen Robotern

“…Ein ähnlicher Multisensoraufbau wurde in [11] benutzt. Dort kommt ein Lasersensor zum Einsatz, welcher sowohl Tiefen-als auch Grauwertdaten liefert, jedoch nur in einem beschränkten Winkelbereich (60°) vor dem Roboter.…”

“…Wie in [11] werden die Unsicherheitsgrenzen der geschätzten Roboterposition untersucht, die unter der Annahme guter Modelle auch den Rückschluss auf die gesuchten ersten Momente zulassen. Aus diesem Grund wird versucht, physikalisch gut basierte Unsicherheitsmodelle für die eingesetzten Sensoren zu bekommen.…”

Multisensor on-the-fly localization:

“…The combination of these features is used in [16,18]. In [16], a laser sensor with an opening angle of 60 • providing both range and intensity images was utilized, and in a recent work [18], the absolute localization accuracy of laser, monocular and trinocular vision was examined. Similar precision has been found for the three cases.…”

Sensor influence in the performance of simultaneous mobile robot localization and map building

“…In our feature-based approach, uncertain geometric information is represented using a probabilistic model: the Symmetries and Perturbation Model (SPmodel) 4,5] which combines the use of probability theory to represent the imprecision in the location of a geometric element, and the theory of symmetries to represent the partiallity due to characteristics of each t ype of geometric element.…”

“…In our approach w e use a probabilistic model, the Symmetries and Perturbation Model (SPmodel) 4,5], to represent uncertain geometric information given by a n y sensor in a general and systematic way. As reported in the literature, the solution to the simultaneous robot localization and map building problem requires maintaining a representation of the relationships between the location estimations of the robot and the features included in the map 6,7], which in our work are represented by the cross-correlations between their estimations.…”