A SEQUENTIAL FACTORIZATION METHOD FOR EUCLIDEAN RECONSTRUCTION FROM IMAGE SEQUENCES WITH MISSING DATA(International Workshop on Advanced Image Technology 2007)

スポンサーリンク

概要

• 論文の詳細を見る

• In this paper we propose a sequential factorization method for solving the structure from motion problem. The proposed method integrates partial reconstruction estimated at each frame for obtaining full reconstruction. It can deal with missing data and treats past information on the camera motion in efficient manner by using principal component analysis (PCA), which reduces computational cost. Our proposed method enables accurate estimation of 3D scene structure and camera motion by using projective camera model. Additionally, the proposed method uses only linear computation, and does not require an initial estimation in contrast to non-linear method. The experimental result on real data shows that this method is reliable under practical situations.

• 社団法人電子情報通信学会の論文
• 2007-01-03

著者

• Mitsuhashi Wataru

  The University Of Electro-communications

• Shibusawa Eijiro

  The University of Electro-Communications

• Shibusawa Eijiro

  Univ. Electro‐communications Tokyo Jpn

関連論文

• ITERATIVE FACTORIZATION APPROACH TO PROJECTIVE RECONSTRUCTION FROM UNCALIBRATED IMAGES WITH OCCLUSIONS(International Workshop on Advanced Image Technology 2009)
• ROBUST PROJECTIVE FACTORIZATION IN THE PRESENCE OF MISSING OR UNCERTAIN DATA(INTERNATIONAL Workshop on Advanced Image Technology 2008)
• A SEQUENTIAL FACTORIZATION METHOD FOR EUCLIDEAN RECONSTRUCTION FROM IMAGE SEQUENCES WITH MISSING DATA(International Workshop on Advanced Image Technology 2007)

スポンサーリンク