A study on accurate 3-D reconstruction from images under restricted condition and environment

2017 Fiscal Year Final Research Report

• Project/Area Number: 15K00232
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Perceptual information processing
• Research Institution: Toyohashi University of Technology
• Principal Investigator: KANAZAWA Yasushi 豊橋技術科学大学, 工学(系)研究科(研究院), 准教授 (50214432)
• Co-Investigator(Kenkyū-buntansha): 太田 直哉 群馬大学, 大学院理工学府, 教授 (10270860)
• Project Period (FY): 2015-04-01 – 2018-03-31
• Keywords: 3次元復元 / 内視鏡画像 / ドローン / 砂浜形状

Outline of Final Research Achievements

The purpose of this research is to reconstruct an accurate 3-D shape of an environment from images taken with the camera under a restricted condition and environment. For doing this, we have proposed an accurate image matching method and accurate parameter computation methods. In 3-D reconstruction from the endoscopic camera image, we showed the effectiveness of our method that adopt a frame selection method using "shape cost" and a focal length estimation method using Gibbs sampling. In 3-D reconstruction from UAV camera image, we also showed that we can obtain very accurate 3-D reconstruction of sands by using homography-based reconstruction and proposed frame selection method. We have realize real-time 3-D reconstruction system from a camera of UAV.

Free Research Field

画像処理・コンピュータビジョン