A study on image matching and description without knowledges about a scene

2014 Fiscal Year Final Research Report

• Project/Area Number: 24500202
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Perception information processing/Intelligent robotics
• 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): OHTA Naoya 群馬大学, 理工学研究院, 教授 (10270860)
• Project Period (FY): 2012-04-01 – 2015-03-31
• Keywords: 画像の対応付け / 特徴量

Outline of Final Research Achievements

We have presented new approach for image matching. In image matching, we usually choose a method, which is famous one or is good from own experience. However, the chosen method is not optimal. Here, we proposed a method that adaptively and automatically selects ``good'' feature extraction and description method. In our method, we first divide an image into small regions and then adopt good feature extractor and descriptor according to their characteristics. We can improve the accuracy of image matching by our approach. While there are some problems in segmentation and matching computation, we have shown the effectiveness of our method.

Free Research Field

コンピュータビジョン・画像処理