Results 1 - 10
of
11
Parametric Ego-Motion Estimation for Vehicle Surround Analysis Using Omni-Directional Camera
, 2004
"... Omni-directional cameras which give 360 degree panoramic view of the surroundings have recently been used in many applications such as robotics, navigation and surveillance. This paper describes the application of parametric ego-motion estimation for vehicle detection to perform surround analysis u ..."
Abstract
-
Cited by 28 (12 self)
- Add to MetaCart
Omni-directional cameras which give 360 degree panoramic view of the surroundings have recently been used in many applications such as robotics, navigation and surveillance. This paper describes the application of parametric ego-motion estimation for vehicle detection to perform surround analysis using an automobile mounted camera. For this purpose, the parametric planar motion model is integrated with the transformations to compensate distortion in omni-directional images. The framework is used to detect objects with independent motion or height above the road. Camera calibration as well as the approximate vehicle speed obtained from CAN bus are integrated with the motion information from spatial and temporal gradients using Bayesian approach. The approach is tested for various configurations of automobile mounted omni camera as well as rectilinear camera. Successful detection and tracking of moving vehicles, and generation of surround map is demonstrated for application to intelligent driver support. Key words Motion estimation, Panoramic vision, Intelligent vehicles, Driver support systems, Collision avoidance 1
Motion analysis of omni-directional video streams for a mobile sentry
- ACM International Workshop on Video Surveillance
, 2003
"... A mobile platform mounted with Omni-Directional Vision Sensor (ODVS) can be used to monitor large areas and detect interesting events such as independently moving persons and vehicles. To avoid false alarms due to extraneous features, the image motion induced by the moving platform should be compens ..."
Abstract
-
Cited by 7 (3 self)
- Add to MetaCart
(Show Context)
A mobile platform mounted with Omni-Directional Vision Sensor (ODVS) can be used to monitor large areas and detect interesting events such as independently moving persons and vehicles. To avoid false alarms due to extraneous features, the image motion induced by the moving platform should be compensated. This paper describes a formulation of parametric ego-motion compensation for an ODVS. Omni images give 360 degrees view of surroundings but undergo considerable image distortion. To account for these distortions, the parametric planar motion model is integrated with the transformations into omni image space. Prior knowledge of approximate camera calibration and vehicle speed are integrated with the estimation process using Bayesian approach. Iterative, coarse to fine, gradient based estimation is used to correct the motion parameters for vibrations and other inaccuracies in prior knowledge. Experiments with camera mounted on a mobile platform demonstrate successful detection of moving persons and vehicles.
Fast camera motion estimation for hand-held devices and applications
- in Proc. 4th Int. Conf. Mobile and Ubiquitous Multimedia
, 2005
"... In this paper we present an efficient motion estimation algorithm for camera-enabled handheld devices, such as mobile phones and PDAs. Compared to general camera motion estimation, the estimation of ego-motion of handheld devices presents unique challenges because most devices are limited in resourc ..."
Abstract
-
Cited by 7 (2 self)
- Add to MetaCart
(Show Context)
In this paper we present an efficient motion estimation algorithm for camera-enabled handheld devices, such as mobile phones and PDAs. Compared to general camera motion estimation, the estimation of ego-motion of handheld devices presents unique challenges because most devices are limited in resources (processing power, memory and battery). The algorithm must be lightweight so it can be efficiently embedded and run fast enough to produce smooth seamless translation. Our solution includes a multi-resolution scheme which searches the match from coarse to fine; and optimization of the search space. As a demonstration, we implemented the algorithm on the Symbian based Nokia 3650/6600 camera phone, and explored several interesting applications, including using camera motion for browsing documents and as a pointing device. A cross-platform implementation is also discussed.
Motion Based Vehicle Surround Analysis Using an Omni-Directional Camera
, 2004
"... Omni-directional cameras which give 360 degree panoramic view of the surroundings have recently been used in many applications such as robotics, navigation and surveillance. This paper describes the application of motion estimation on omni camera to perform surround analysis using an automobile moun ..."
Abstract
-
Cited by 5 (2 self)
- Add to MetaCart
Omni-directional cameras which give 360 degree panoramic view of the surroundings have recently been used in many applications such as robotics, navigation and surveillance. This paper describes the application of motion estimation on omni camera to perform surround analysis using an automobile mounted camera. The system detects and tracks the surrounding vehicles by compensating the ego-motion and detecting objects having independent motion. Prior knowledge about ego-motion and calibration is optimally combined with the information from the image gradients to get better motion compensation.
Motion Analysis for Event Detection and Tracking with a Mobile Omni-Directional Camera
, 2004
"... A mobile platform mounted with Omni-Directional Vision Sensor (ODVS) can be used to monitor large areas and detect interesting events such as independently moving persons and vehicles. To avoid false alarms due to extraneous features, the image motion induced by the moving platform should be compens ..."
Abstract
-
Cited by 3 (0 self)
- Add to MetaCart
A mobile platform mounted with Omni-Directional Vision Sensor (ODVS) can be used to monitor large areas and detect interesting events such as independently moving persons and vehicles. To avoid false alarms due to extraneous features, the image motion induced by the moving platform should be compensated. This paper describes a formulation and application of parametric ego-motion compensation for an ODVS. Omni images give 360 degrees view of surroundings but undergo considerable image distortion. To account for these distortions, the parametric planar motion model is integrated with the transformations into omni image space. Prior knowledge of approximate camera calibration and camera speed are integrated with the estimation process using Bayesian approach. Iterative, coarse to fine, gradient based estimation is used to correct the motion parameters for vibrations and other inaccuracies in prior knowledge. Experiments with camera mounted on various types of mobile platforms demonstrate successful detection of moving persons and vehicles.
Structure from Small Baseline Motion with Central Panoramic Cameras
- IEEE WORKSHOP ON OMNIDIRECTIONAL VISION (IN CONJUNCTION WITH CVPR
, 2003
"... In applications of egomotion estimation, such as real-time vision-based navigation, one must deal with the double-edged sword of small relative motions between images. On one hand, tracking feature points is easier, while on the other, two-view structure-from-motion algorithms are poorly conditioned ..."
Abstract
-
Cited by 3 (0 self)
- Add to MetaCart
In applications of egomotion estimation, such as real-time vision-based navigation, one must deal with the double-edged sword of small relative motions between images. On one hand, tracking feature points is easier, while on the other, two-view structure-from-motion algorithms are poorly conditioned due to the low signal-to-noise ratio. In this paper, we derive a multi-frame structure from motion algorithm for calibrated central panoramic cameras. Our algorithm avoids the conditioning problem by explicitly incorporating the small baseline assumption in the algorithm 's design. The proposed algorithm is linear, amenable to real-time implementation, and performs well in the small baseline domain for which it is designed.
Article Odometry for Ground Moving Agents by Optic Flow Recorded with Optical Mouse Chips
, 2014
"... sensors ..."
Towards removing artificial landmarks for autonomous exploration in structured environments
, 2005
"... Abstract — This research takes one step towards using naturally-occurring and automatically detected features present in a structured environment for map building. We present a simple real-time algorithm for solving the vision task of extracting interesting cues from the environment, to be used as l ..."
Abstract
-
Cited by 1 (0 self)
- Add to MetaCart
(Show Context)
Abstract — This research takes one step towards using naturally-occurring and automatically detected features present in a structured environment for map building. We present a simple real-time algorithm for solving the vision task of extracting interesting cues from the environment, to be used as landmarks. The landmarks are used in a simultaneous localization and mapping (SLAM) algorithm based on an extended Kalman filter. The natural feature based map is embedded in a topological framework to map a large office environment on the order of 4000 m 2. This map is later used for global localization of the robot. Experimental results on a mobile robot are included. We outline a philosophy of good landmark while deriving an analytical expression of sparseness between two landmarks for good map building.
Fuzzy edge detection for omnidirectional images
, 2008
"... The use of omnidirectional vision has increased during these past years. It provides a very large field of view. Nevertheless, omnidirectional images contain significant radial distortions and conventional image processing is not adapted to these specific images. This paper presents an edge detector ..."
Abstract
-
Cited by 1 (0 self)
- Add to MetaCart
The use of omnidirectional vision has increased during these past years. It provides a very large field of view. Nevertheless, omnidirectional images contain significant radial distortions and conventional image processing is not adapted to these specific images. This paper presents an edge detector adapted to the image geometry. Fuzzy sets will be used to take into account all imprecisions introduced by the sampling process. The Prewitt filter applied to omnidirectional image will be studied to illustrate this paper.
nn a Rotation estimation Translation estimation Motion decoupling
"... n t f v tion ral c res it c by decoupling the rotation and the translation. The rotation is computed by an efficient algorithm based the s e of th n cont ot, sev when GPS or inertial guidance is not available [31,48,16,60]. Most of the existing works use conventional cameras that have a relative sma ..."
Abstract
- Add to MetaCart
(Show Context)
n t f v tion ral c res it c by decoupling the rotation and the translation. The rotation is computed by an efficient algorithm based the s e of th n cont ot, sev when GPS or inertial guidance is not available [31,48,16,60]. Most of the existing works use conventional cameras that have a relative small field of view, which might lead to important difficulties (lack of features, translation/rotation ambiguity, etc.) Intuitively, wider the field of view is, the more information we can gather from the environment (Fig. 1). Therefore, an imaging system that is able to estimation problem which has been studied for several applica-tions such as visual odometry [9], SLAM [34] or structure from mo-tion [22,36]. In this paper, we aim to demonstrate the efficiency of our approach for the motion estimation problem with respect to three criteria: accuracy, robustness and speed. Accuracy refers to the precision of the rotation and translation estimations. For the robustness part, we study the dependency of our method to noised matching points and inaccurate rotations. For the speed part, the complexity of the proposed methods is analyzed. Before presenting our approach, we review the existingmethods of motion estimation. They can be classified into three categories: * Corresponding author.
