Catadioptric stereo using planar mirrors.

详细信息   

• 作者：Gluckman ; Joshua Michael.
• 学历：Doctor
• 年：2000
• 导师：Nayar, Shree K.
• 毕业院校：Columbia University
• 专业：Computer Science.
• ISBN：9780599923966
• CBH：9985891
• Country：USA
• 语种：English
• FileSize：4029562
• Pages：133

文摘

By using mirror reflections of a scene, stereo images can be captured with a single camera (catadioptric stereo). In addition to simplifying data acquisition single camera stereo provides both geometric and radiometric advantages over traditional two camera stereo. Although several catadioptric stereo sensors have been proposed there has been no study of the benefits and limitations of such systems.;In the first part of this thesis, we discuss the geometry and calibration of catadioptric stereo with planar mirrors. In particular, we examine the relative orientation, the epipolar geometry and the possibility of self-calibration when using mirrors placed in an arbitrary configuration.;One of the benefits of monocular stereo is the quality of the stereo data that is obtained. When rectifying stereo images it is important to maintain this quality. In the second part, we present a novel method for stereo rectification that minimizes the loss of information due to resampling. First, we describe a metric that quantifies the information loss when a perspective transformation is applied. Then we derive the complete class of rectifying perspective transformations and show how to find the transformation within this class, that minimizes the loss of information.;In the third part of this thesis, we present novel catadioptric sensors which use mirrors to produce optically rectified stereo images. The scan-line correspondence of these images benefits real-time stereo by avoiding the computational cost and image degradation due to resampling when rectification is performed after image capture. First, we develop a theory which determines the number of mirrors that must be used and the constraints on those mirrors that must be satisfied to obtain rectified stereo images with a single camera. Then we discuss in detail the use of both one and three mirrors and analyze the effect of misplacement errors. In addition, we show how the mirrors should be configured in order to minimize sensor size for a given baseline, an important design consideration. Finally, we will describe a real-time system which demonstrates the viability of stereo with mirrors as an alternative to traditional two camera stereo.