A statistical framework for long-range feature matching in uncalibrated image mosaicing

被引：18

作者：

Cham, TJ ^{[1
]}

Cipolla, R ^{[1
]}

机构：

[1] Digital Equipment Corp, Cambridge Res Lab, Cambridge, MA 02139 USA

来源：

1998 IEEE COMPUTER SOCIETY CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, PROCEEDINGS | 1998年

关键词：

D O I：

10.1109/CVPR.1998.698643

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

The problem considered in this paper is that of estimating the projective transformation between two images in situations where the image motion is large and feature-matching is nor aided by a proximity heuristic. The overall algorithm designed is based on a multiresolution, multi-hypothesis scheme, and similarities between tracking and matching through multiple resolution levels are exploited. Two major tools are developed in this paper (i) a Bayesian framework for incorporating similarity measures of feature correspondences in regression to specify the different levels of confidence in the correspondences; and (ii) a Bayesian version of RANSAC, which is able to utilise prior estimates and matching probabilities. The algorithm is tested on a number of real images with large image motion and promising results were obtained.

引用

页码：442 / 447

页数：6