D.J. Kriegmen, E. Triendl and T.O. Binford. Stereo vision and navigation in buildings for mobile robots. IEEE Trans. on Robotics and Automation, 5(6):1722--1727, 1991.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....conditional probability distributions of each estimate: y 1 , y 2 ) p(y 1 x)p(y 2 x)p(x) 1) The combined estimate x is the maximum a posteriori estimate of x obtained from this distribution. Examples of frameworks for cue integration that are defined this way include Kalman filters [5, 6] and physically based frameworks [15, 16] These probabilistic approaches are demonstrated pictorially in Figure 2(b) showing how the combined estimate x weights together the individual results based on their reliabilities, and produces an estimate which has less uncertainty (depicted as an ....

Kriegman, D., Triendl, E., Binford, T.: Stereo vision and navigation in buildings for mobile robots. IEEE Transactions on Robotics and Automation 5 (1989) 792--803

....sensing uncertainty except that uncertainties are represented by probabilistic distributions. However, a concrete model of sensors is needed to solve a planning problem of an actual robot. We here describe a model of stereo vision as an example of modeling uncertainty. Although most researches [1][6][7] deal with only quantization error in edge detection, uncertainty caused by correspondence of wrong pairs of features also needs to be investigated because this uncertainty is much greater than quantization errors. A. Quantization Error in Edge Detection Suppose a point at (x, z) be observed ....

....errors. A. Quantization Error in Edge Detection Suppose a point at (x, z) be observed at XL and XR on the left and the right images, respectively. We assume that the real positions of XL and XR are normally and independently distributed, and linearize the equation of the image projection [6]. Then, the uncertainty of (x, z) can be represented by a two dimensional normal distribution. Although the distribution of (XL,X R ) depends on the contrast of each edge, we currently assume that this distribution always has the same covariance matrix. Using this model, distributions of other ....

Kriegman, D.J., Triendl, E. and Binford, T.O., "Stereo Vision and Navigation in Buildings for Mobile Robots," IEEE Trans. Robotics and Automat., vol.5, no.6, pp.792803 (1989).

....new measurements. This is similar to the predict and Figure 1: The two obstacles in the range image on the left are mapped into the planar view on the right. A column in the range image is condensed into two values in the planar view: the minimum and maximum range. match methods of Kriegman et al. [KTB89] which used line segments instead of regions. Range shadows are explicitly modeled as continuous areas between measurements. The angular representation leaves the uncertainty in the measurements along the radial direction and does not mix the uncertainty with the angular resolution which is quite ....

....are updated from frame to frame based on the uncertainty in each measurement. The minimum and maximum range values are updated based on the relative uncertainties of the new measurement and the current model. This is similar to recursive estimation via Kalman filtering. A similar example is [KTB89]. The longer an object is tracked in the image, the smaller the uncertainty in model parameters. Those objects which have small enough uncertainty are added to a global map when the vehicle passes them. This global map can be used to identify landmarks for localization. 4 Experiments We ran the ....

D.J. Kriegman, E. Triendl, and T.O. Binford. Stereo vision and navigation in buildings for mobile robots. IEEE Robotics and Automation, 5(6), 1989.

.... Weighted averaging method takes a weighted average of the composite signals, where the weights are determined by the estimated variances of the signals [20] Kalman filtering uses the statistical characteristics of a measurement model to recursively determine estimates for the fused data [57]. 2.2.2.2 Pixel level fusion Pixel level fusion can be used to increase the information content associated with each pixel in an image formed through a combination of multiple images, thus it may result in more reliable segmentation and more discriminating features for further processing. ....

D. J. Kriegman, T. E., and T. O. Binford, "Stereo vision and navigation in buildings for mobile robots," IEEE Trans. Robot. Automat., vol. 5, no. 6, pp. 792--803, 1989.

....indoor mobile robots seek quantitative, numeric answers to questions regarding position esD R A F T November 30, 1999, 7:52pm D R A F T 6 timates. The search for the best position estimate can be conducted in two different spaces: 1) the space of correspondences between image and map features [2, 24, 25], or (2) the space of position estimates [7, 8, 21, 40, 49] The search in the space of position estimates is often conducted by quantizing this space into a two dimensional array, an approach inspired by the occupancy grids developed for map making [13, 14, 34, 35] Position can also be ....

D. J. Kriegman, E. Triendl, and T. O. Binford. Stereo vision and navigation in buildings for mobile robots. IEEE Transactions on Robotics and Automation, 5(6), December 1989.

....angles. This representation is called the sensed sphere [51] The sensed sphere can be constructed either by a laser or by a stereo approach. To use a stereo approach to construct the sensed sphere, we can first construct a sparse depth sampling of the environment by existing stereo algorithms [32, 38, 48, 54] and then transform this into the representation of a sensed sphere. In this paper, we used a laser range finder to construct the sensed sphere. In order to do this, we need to tessellate the surface of the unit sphere so as to get a series of uniform or near uniform patches which partition the ....

D. J. Kriegman, E. Triendl, and T. O. Binford, Stereo vision and navigation in building for mobile robots, IEEE Trans. Robotics Automat. 5(6), 1989, 792--803.

....more common for fusing information. In [18] stereo and shading information are combined using a soft constraint (weighted terms from each source are added into the system energy) A physics based sensor fusion method combining range and intensity data was presented in [51] Using a Kalman filter [2, 26] or Bayesian methods [12] for fusion combines solutions in a similar way. Aside from this, Kalman filtering has become a standard tool for estimation in dealing with noisy data [3, 19, 30, 32, 37] Face tracking: There is a vast body of work on tracking the human face, with applications ranging ....

....the use of a Kalman filter. This section describes how the computation from Section 4.2 is reformulated using an iterated extended Kalman filter. Kalman filtering [3, 19, 30] has become a popular tool in computer vision, and the formulation here is, on the whole, similar to other applications [2, 7, 26, 32, 37]: there is a measurement equation which models the noise inherent in the data gathering process, and there is a process model, which predicts the behavior of the system based on the current state. The initialization and tuning of the filter is accomplished using standard techniques. The ....

[Article contains additional citation context not shown here]

D.J. Kriegman, E. Triendl, and T.O. Binford. Stereo vision and navigation in buildings for mobile robots. RA, 5(6):792--803, December 1989. 40

....algorithm under different constraints on motion. Speedups of up to 400 are achieved without significant errors. 1 Introduction Stereo vision is one of the most common and robust vision algorithms used in mobile robot navigation. It has been used for mapping, localization and obstacle avoidance [8], 10] 11] While performance of stereo has increased with growing computer power, the technique is still limited by the high computational cost of the algorithm. One reason for the lack of performance is that stereo is performed fully in each iteration of the perceptual cycle. This is wasteful ....

E. Kreigman, D.J. Triendl and T.O. Binford. Stereo vision and navigation in buildings for mobile robots. IEEE Transactions on Robotics and Automation, pages 792--803, June 18-20 1989.

....1 This work was supported in part by DARPA and TACOM under contract DAAE07 91 C R035 1. INTRODUCTION Robot navigation has been the focus of attention of many researchers in recent years. Some of the systems need to be very accurately calibrated in order to perform navigation successfully[5, 3]. Calibration is the process of determining the relationship of sensor output to the actual value of the input. In the system described in [3] it was necessary to position the robot camera system by hand to within 0.1 inches and 0.1 ffi of a known position in a world coordinate system. This ....

D. Kriegman, E. Triendl, and T. Binford. Stereo vision and navigation in buildings for mobile robots. Robotics and Automation, 5(6), 1989.

....of the camera. This work lies in the last category, with a 3D environment and a calibrated camera. The existence of vertical edges in the scene supports the building of a 2D map in the first step of our algorithm. Similar 2Dlay out computations can be found in the works of Brooks [6] Kriegman [15] Kak [14] Lebegue [17] and Betke [4] It should be mentioned here that the vertical line formulation is equivalent with using 1D projections of points (a horizontal slice of the environment) in order to recover a 2D map. Faugeras et al. 9] formulated and solved the problem of 2D recovery from ....

D.J. Kriegman, E. Triendl, and T.O. Binford. Stereo vision and navigation in buildings for mobile robots. Trans. on Robotics and Automation, 5:792--804, 1989.

....into the distance between the sensor and the nearest obstacle in the view of the sensor. Visual information can also be used to gather information about the surroundings of a mobile robot. Structured lighting or stereo vision, for example, can provide range information about the environment [8]. There are two broad categories of environmental representation: grid based and feature based. In a grid based representation, the environment is segmented into squares 1 . Based on measurements of the environment and a priori information, each square is assigned a value of empty or full. It is ....

D. Kriegman, E. Triendl, and T. Binford. Stereo vision and navigation in buildings for mobile robots. IEEE Transactions on Robotics and Automation, 5(6), December 1989.

....to support both accurate or qualitative navigation. Results are described in the paper. 1. Introduction In recent years visually guided navigation [19] has been approached in many different ways. Map building using image sequences from standard single or stereo cameras has been explored in [2, 20]. Kosaka and Kak [18] combined model based reasoning and Kalman filtering, which requires a 3D geometric model of the environment. In [17] sonar data is collected along corridors and used together with visual data (simply vertical edges) to aid navigation. Although previous research produced many ....

D. Kriegman, E. Triendl, and T. Binford. Stereo vision and navigation in buildings for mobile robots. IEEE Transactions on Robotics and Automation, 5(6):792--802, December 1989.

....the quaternion algebra, a method which has recently regained popularity, particularly in the area of computer vision. Most recent work in autonomous robot navigation has been done in structured indoor environments. A sampling of this work can be found in [Crowley, 1985, Tsubouchi and Yuta, 1987, Kriegman et al. 1989, Kosaka and Kak, 1992, Fennema et al. 1990, Ferrari et al. 1990, Atiya and Hager, 1993, D Orazio et al. 1992] Of the work in outdoor navigation, the following is associated most closely with this research. Dickmanns experimental five ton van VaMoRs navigated in outdoor environments, but ....

D. J. Kriegman, E. Triendl, and T. O. Binford. Stereo vision and navigation in buildings for mobile robots. IEEE Transactions on Robotics and Automation, 5(6):792--803, October 1989.

....technique: are those methods where only the relative motion of the vehicle is estimated (rotation and translation) so for goal oriented applications the absolute location of the start point must be known. Those techniques, however, do not require any a priori knowledge about the environment [6, 1, 2]. Methods in both classes work with a good performances (referred to execution time) and provide an acceptable estimation of the vehicle location and orientation. Most of them, however, have a fundamental drawback: they do not provide an estimation of the uncertainty in the vehicle location and ....

O.Binford D.J.Kriegman, E.Triendl. Stereo vision and navigation in buildings for mobile robots. IEEE trans. on Robotics and Automation, 5(6):792--803, 1989.

....poses no experimental difficulty one simply takes the measured angles and applies the appropriate transformation to them before averaging. 1.1. 5 Related Work There are several works from the eighties early nineties which perform reconstruction using vertical edges and known motion (Kriegman [7], Kak [6] There is work in structure from motion from circular trajectories (Shariat and Price [12] Sawhney [11] This work, however, uses a set of equations with the constraint that the projection centers are on a circle. Recently, a number of approaches to navigation and reconstruction using ....

D.J. Kriegman, E. Triendl, and T.O. Binford. Stereo vision and navigation in buildings for mobile robots. Trans. on Robotics and Automation, 5:792--804, 1989.

....in order to obtain an accurate position estimate. Extended Kalman filtering techniques have been used in a variety of robotic tasks such as road following, visual map building and egomotion estimation and other related navigational applications such as ship navigation and missile tracking [19, 3, 33, 44, 25]. One significant disadvantage of both the Kalman filter and the EKF is that they make a locally linear approximation to the true relationship between position and observations. That is, the Kalman filtering techniques tend to rely on a good a priori estimate. For this reason Kalman filtering ....

David J. Kriegman, Ernst Triendl, and Thomas O. Binford, Stereo vision and navigation in buildings for mobile robots, IEEE Transactions on Robotics and Automation 5 (1989), no. 6, 792--803.

....presented in this paper is that they do not require a global positioning system, which simplifies their implementation. Most of the research that has been reported in the robotics literature casts the exploration problem in terms of constructing a global metric map of the robot s workspace [10], 1] 14] If the robot were able to construct such a map, then it could employ classical path planning algorithms [3] 16] to navigate through its environment. Two systematic techniques (named the Sightseer and Seed Spreader strategies) for exploring the geometry of a two dimensional ....

D. J. Kriegman, E. Triendl, and T. O. Binford. Stereo vision and navigation in buildings for mobile robots. IEEE Trans. on Robotics and Automation, 5(6):792--803, December 1989.

No context found.

D.J. Kriegmen, E. Triendl and T.O. Binford. Stereo vision and navigation in buildings for mobile robots. IEEE Trans. on Robotics and Automation, 5(6):1722--1727, 1991.

No context found.

D.J. Kriegmen, E. Triendl, and T.O. Binford, Stereo vision and navigation in buildings for mobile robots, IEEE Trans. on Robotics and Automation,5(6):1722-1727, 1991.

No context found.

Kriegman D. J., Triendl E. and Binford T. O. (1989), Stereo vision and navigation in buildings for mobile robots IEEE Transactions on Robotics and Automation 5(6) pp. 792-803

No context found.

D. Kriegman, E. Triendl, and T. Binford. Stereo vision and navigation in buildings for mobile robots. IEEE Trans. Robotics and Automation, 5(6), December 1989.

No context found.

D. J. Kriegman, E. Triendl, and T. O. Binford, "Stereo vision and navigation in buildings for mobile robots," IEEE Trans. Robot. Automat., vol. 5, pp. 792--803, Dec. 1989.

No context found.

Kriegman, D.J., Triendl E. and Binford, T.O., "Stereo Vision and Navigation in Buildings for Mobile Robots,"IEEE Trans. on Robotics and Automation, 5(6): 792-803.

No context found.

D.J. Kriegman, E. Triendl, and T.O. Binford. Stereo vision and navigation in buildings for mobile robots. Trans. on Robotics and Automation, 5:792#804, 1989.

No context found.

Kriegmaan, D.J., Triendl, E. and Binford, T.O. (1989). Stereo vision and navigation in buildings for mobile robots. IEEE Trans. Robotics and Automation, 5(6), pp. 792-803.

First 50 documents

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC