The goal of intra-surgical registration is to establish a common reference frame between pre-surgical and intra-surgical 3-D data sets that correspond to the same anatomy. This paper presents two novel techniques which have application to this problem: high-speed pose tracking, and intra-surgical data selection. In the first part of this paper, we describe an approach for tracking the pose of arbitrarily-shaped rigid objects at rates up to 10Hz. Static accuracies on the order of 1mm in translation and 1 degree in rotation have been achieved. We have demonstrated the technique on a human face using a high-speed VLSI range sensor; however, the technique is independent of the sensor used or the anatomy tracked. In the second part of this paper, we describe a general purpose approach for selecting near-optimal, intrasurgical registration data. Due to high costs associated with the acquisition of intra-surgical data, it is desirable to minimize the amount of data acquired, while ensuring that registration accuracy requirements are met. We synthesize near-optimal intra-surgical data sets, based upon an analysis of differential surface properties of pre-surgical data. We demonstrate, using data from a human femur, that discrete point data sets selected using our method provide superior pose refinement accuracy to those selected by

We propose a novel method for combining information streamed by a multi-sensor system for Visual Surveillance. Information fusion occurs in two phases during which all cameras are calibrated with respect to independent global Cartesian reference frames (set on the ground plane) and then all frames are registered into a single coordinate system. The development of automatic calbration and registering of visual data is crucial in Visual Surveillance applications because it makes easier to install the monitoring infrastructure and, consequently, to develop more accessible Visual Surveillance tools for the public domain. Machine Learning techniques are believed to offer the best mathematical tools to handle the uncertainy and incomplete nature of surveillance data. 1

Determination of relative three-dimensional (3--D) position, orientation, and relative motion between two reference frames is an important problem in robotic guidance, manipulation, and assembly as well as in other fields such as photogrammetry. A solution to this problem that uses two-dimensional (2--D), intensity images from a single camera is desirable for real-time applications. Where the object geometry is unknown, the estimation of structure is also required. A single camera is advantageous because a standard video camera is low in cost, setup and calibration are simple, physical space requirements are small, reliability is high, and low-cost hardware is available for digitizing and processing the images. A di#culty in performing this measurement is the process of projecting 3--D object features to 2--D images, a nonlinear transformation. Noise is present in the form of perturbations to the assumed process dynamics, imperfections in system modeling, and errors in the feature locations extracted from the 2--D images. This dissertation presents solutions to the remote measurement problem for a dynamic system given a sequence of 2--D intensity images of an object where feature positions of the object are known relative to a base reference frame and where the feature positions are unknown relative to a base reference frame. The 3--D transformation is modeled as a nonlinear stochastic system with the state estimate providing six degree-of-freedom motion and position values. The stochastic model uses the iterated extended Kalman filter as an estimator and as a screw representation of the 3--D transformation based on dual quaternions. Dual quaternions provide a means to represent both rotation and translation in a unified notation. The method has been implemented and tes...

In order to monitor a robot's position, traditional techniques have used a model of the environment and a 2D view (image) of the scene from the current position of the robot. These techniques require either a fairly accurate 3D scene model or noisy 3D models with a reliable estimate of the model noise. Since it is difficult either to automatically obtain accurate 3D models or to reliably estimate the model noise, the 3D models are constructed manually. In this paper we propose automating the model generation step by using a robust model acquisition algorithm which also provides an estimate of the error in the acquired model. This paper shows that the acquired 3D model along with the error estimate is reliable enough for a position estimation algorithm.

In order to monitor a robot's position, traditional techniques have used a model of the environment and a 2D view (image) of the scene from the current position of the robot. These techniques require either a fairly accurate 3D scene model or noisy 3D models with a reliable estimate of the model noise. Since it has not been possible either to automatically obtain accurate 3D models or to reliably estimate the error in noisy models, the 3D models have been constructed manually in previous work. In this paper we propose automating the model generation step by using a robust model acquisition algorithm. Apart from automatically obtaining fairly accurate 3D scene models, this model acquisition technique also provides the error in the model; this paper shows that the acquired 3D model along with the error estimate is reliable enough for a position estimation algorithm. 1 Introduction Traditionally, a moving robot monitors its position in an environment by using a 3D model of the scene and ...

Given a 3D object and some measurements for points in this object, it is desired to find the 3D location of the object. A new model based pose estimator from stereo pairs based on linear programming (LP) is presented. In the presence of outliers, the new LP estimator provides better results than maximum likelihood estimators such as weighted least squares, and is usually almost as good as robust estimators such as LMEDS. In the presence of noise the new LP estimator provides better results than robust estimators such as LMEDS, and is slightly inferior to maximum likelihood estimators such as weighted least squares. In the presence of noise and outliers - especially for wide angle stereo - the new estimator provides the best results. The LP estimator is based on correspondence of a points to convex polyhedrons. Each points corresponds to a unique polyhedron, which represents its uncertainty in 3D as computed from the stereo pair. Polyhedron can also be computed for 2D data point by usi...