Light Field Vision - Computational Photography for Image Understanding

2016 Fiscal Year Final Research Report

• Project/Area Number: 25240027
• Research Category: Grant-in-Aid for Scientific Research (A)
• Allocation Type: Single-year Grants
• Section: 一般
• Research Field: Perceptual information processing
• Research Institution: Kyushu University
• Principal Investigator: Nagahara Hajime 九州大学, システム情報科学研究院, 准教授 (80362648)
• Co-Investigator(Kenkyū-buntansha): 谷口 倫一郎 九州大学, システム情報科学研究科(研究院, 教授 (20136550) ; 鳥居 秋彦 東京工業大学, 理工学研究科, 助教 (20585179) ; 日浦 慎作 広島市立大学, 情報科学研究科, 教授 (40314405) ; 向川 康博 奈良先端科学技術大学院大学, 情報科学研究科, 教授 (60294435) ; 島田 敬士 九州大学, 学内共同利用施設等, 准教授 (80452811)
• Project Period (FY): 2013-04-01 – 2017-03-31
• Keywords: コンピュテーショナルフォトグラフィ / ライトフィールド / 物体認識

Outline of Final Research Achievements

It is getting attention to recognize objects from a image. There are a lot of researches have been proposed and most of them uses statistical learning and recognition such as bag of features of intensity pattern of the object. However, this kind of approach cannot directly apply to transparent object recognition, since the appearance and the intensity patten of the object is drastically changed by background appearance. So the previous approach have been ignored the transparent objects from the target object. We proposed new approach and methods to handle the transparent objects in object recognition. We proposed light field distortion feature (LFD) for describing the object. The feature describes the distortion of light field cased by refraction of the transparent objects. We used the LFD feature and realized to object classification and segmentation of the objects as the applications.

Free Research Field

コンピュテーショナルフォトグラフィ