双摄象头法基于边缘的分层配准和距离获取

ACQUIRING RANGE DATA IN BINOCULAR 3-D COMPUTER VISION USING EDGEBASED HIERARCHICAL MATCHING

本文采用基于边缘的分层配准方法,给出了一个双摄象头三维视觉系统。用 Frei-Chen 算子提取边缘,梯度值采用非均匀量化并应用预加重技术。图象压缩采用完全四叉树,用 Illiffe 法存贮和检索。距离计算采用简化了的 Trivedi 公式,并用 Householder 变换求解用于距离计算和系统校准的超定方程组。

This paper deals with a binocular 3-D computer vision system based on the hierarchicalmatching of edge features,Frei and Chen operator is used to extract the edge.The averagegradients of an image obtained by two isotropic operators are non-equal quantized andthresholded in an angle.Edge features are extracted after passing a preemphasis transferfunction which can equalize the noise affection.Binary edge images are decomposed into apyramid structure which is stored and searched using Illiffe's location method.Corre-sponding points are used to determine the range data using triangulation based on an im-proved Trivedi's formula.In calibration the authors set the optical axes of the two camerasparallel (?) simplify the calculation.A 3-rd order Householder transform is used to solve thecompatible coupled equations.