3D Imaging System using Aerial 3D Display

2018 Fiscal Year Final Research Report

• Project/Area Number: 16K12426
• Research Category: Grant-in-Aid for Challenging Exploratory Research
• Allocation Type: Multi-year Fund
• Research Field: Multimedia database
• Research Institution: Tohoku University
• Principal Investigator: Aoki Terumasa 東北大学, 未来科学技術共同研究センター, 准教授 (00302787)
• Project Period (FY): 2016-04-01 – 2019-03-31
• Keywords: Aerial 3D Display / 画像認識 / 3次元点描画 / 形状認識 / 深層学習

Outline of Final Research Achievements

An image recognition method for digital devices (such as smartphones) to recognize 3D point clouds projected by Aerial 3D Display is developed in this research. The core technology to achieve this is development of a novel image recognition method for light-emitting, point-drawing and semi-transparent objects. To solve this problems, two approaches - a shape-recognition based method and deep-learning based method - are developed. And the effectiveness of these methods are shown in this report.

Free Research Field

画像処理・画像理解

Academic Significance and Societal Importance of the Research Achievements

本研究の学術的特色は, (1)発光体, (2)点描画, (3)半透明（点描画の点と点の隙間に背景が写るため）という従来画像処理／画像理解分野で研究対象とされることの少なかった被写体を対象に新しい画像マッチング技術を構築する点である. 本研究の成果はAerial 3D Displayで生成された点描画コンテンツのみならず, 従来認識が困難であった(1)～(3)の性質を持つ被写体に対しても認識を可能にするため, その用途も非常に広い.