Bio-inspired vision system with unified detecting schemes of motion, 3-D, and image features

2014 Fiscal Year Final Research Report

• Project/Area Number: 24300070
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Partial Multi-year Fund
• Section: 一般
• Research Field: Perception information processing/Intelligent robotics
• Research Institution: The University of Tokyo
• Principal Investigator: ANDO Shigeru 東京大学, 情報理工学(系)研究科, 教授 (70134468)
• Project Period (FY): 2012-04-01 – 2015-03-31
• Keywords: 時間相関イメージセンサ / 荷重積分法 / オプティカルフロー / 固視微動 / 特徴抽出 / 超解像 / ステレオ

Outline of Final Research Achievements

Aims of this study are the realization of novel vision systems and algorithms
with outstanding performance based on the bio-inspired sensing principle and the correlation image sensor. The results obtained are: 1) Mathematical formulations of bibroocular detecting scheme obtained using the Gauss and Stokes theorem, the complex differentials, and the optical flow PDE. 2) Image feature extraction algorithms using the 1st and 2nd complex differentials were obtained and a realtime image analysis system using them was implemented. 3) Optical flow detection algorithm based on the phase-stamp imaging was established and a realtime image analysis system was developed. 4) Super resolution principle using the vibroocular imaging was established
and a realtime imaging system was developed. 5) Binocular stereo vision system was constructed and the correspondence detection algorithm between synchronized change of binocular images was established.

Free Research Field

計測，センサ，画像処理，信号処理