Conditional localization and mapping using stereo camera

Authors:
Jigang Liu;Maylor Karhang Leung;Daming Shi
Affiliations:
School of Computer Engineering, Nanyang Technological University, Singapore and School of Electrical Engineering and Computer Science, Kyungpook National University, Daegu, Korea;School of Computer Engineering, Nanyang Technological University, Singapore and School of Electrical Engineering and Computer Science, Kyungpook National University, Daegu, Korea;School of Computer Engineering, Nanyang Technological University, Singapore and School of Electrical Engineering and Computer Science, Kyungpook National University, Daegu, Korea
Venue:
PRICAI'10 Proceedings of the 11th Pacific Rim international conference on Trends in artificial intelligence
Year:
2010

Citing 6
Cited 0

Estimating 3-D rigid body transformations: a comparison of four major algorithms

Machine Vision and Applications - Special issue on performance evaluation
Simultaneous Localization and Map-Building Using Active Vision

IEEE Transactions on Pattern Analysis and Machine Intelligence
MonoSLAM: Real-Time Single Camera SLAM

IEEE Transactions on Pattern Analysis and Machine Intelligence
Vision-Based SLAM: Stereo and Monocular Approaches

International Journal of Computer Vision
Inverse Depth Parametrization for Monocular SLAM

IEEE Transactions on Robotics
Conditional filters for image sequence-based tracking - application to point tracking

IEEE Transactions on Image Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper, conditional localization and mapping (CLAM) is realized with a stereo camera as the only sensor. Compared with visual simultaneous localization and mapping (SLAM), the framework of CLAM is a novel proposed condtional filter rather than extended Kalman filter (EKF). In this algorithm, there is no camera velocity information in the filter state, the measurements and state equation all depend on image data which are the most reliable information so that CLAM outperforms SLAM when the camera turns abruptly or there are some frames lost in which conditions the SLAM may diverge quickly because the predefined model is incorrect in such cases. For CLAM, the model is derived from image data so that CLAM has no such problems. The experimental results show that the proposed CLAM is robust to abrupt turning of the camera and frame-losing, and also give the precise 3D information about the features and the trajectory of the camera.