3D Accuracy Improvement Using an Uncalibrated Image

スポンサーリンク

概要

• 論文の詳細を見る

• In this paper, we focus on a simple, yet important problem: given the coordinates of 3D points and their uncertainties, how much can we improve them provided a new set of corresponding image points. It is difficult, because 3D points, image points and camera parameters and also their uncertainties are closely coupled to each other. These uncertainties are fully modeled and updated utilizing constraints found in the projectivity between 3D world and 2D image plane. The updated, improved uncertainties, represented as covariance matrices, can be used as a "goodness" measure of reconstructed 3D points. All systems which recover 3D scene information, such as active vision systems, mobile robot navigation systems, can profit using this approach. Experimental results with synthesized and real data are shown.

• 一般社団法人電子情報通信学会の論文
• 2002-02-01

著者

• Kinoshita Keisuke

  Atr Human Information Processing Research Laboratories

• TONKO Martin

  ATR Human Information Processing Research Laboratories

関連論文

• Face tracking system in a dynamic environment
• Mouth Tracking from Video Sequences using Trainable Multivariate Gaussian Classifiers
• Mouth Tracking from Video Sequences using Trainable Multivariate Gaussian Classifiers
• 3D Accuracy Improvement Using an Uncalibrated Image

スポンサーリンク