Realtime Color Segmentation, Optical Flow, Depthmap Generation Integration for Mobile Robot Control

用于移动机器人控制的实时颜色分割、光流、深度图生成集成

• 批准号: 10450096
• 负责人: INOUE Hirochika
• 金额: $ 7.74万
• 依托单位: The University of Tokyo
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for Scientific Research (B)
• 财政年份: 1998
• 资助国家: 日本
• 起止时间: 1998 至 1999
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/grant/KAKENHI-PROJECT-10450096/
• 关键词: Real-time vision; Optical flow generation; Depth map generation; Correlation hardware; Mobile robot navigation; 高速両眼カメラヘッド; 相関演算ハードウエア

Actually, from the point of view of safety, working space of robot is separated from life space of human being. In order to utilize a robot in a human life, sensing technology to detect human being is required. This research aims to develop a robot vision system that can detect walking or moving human being while robot is also moving, by combining color segmentation, optical flow generation and depth map generation. Following is a list of research topics,1. Color segmentation method from correlation calculation,2. Boundary detection using optical flow generation,3. Depth map generation from stereo cameras,4. Development of high speed three axis robot head,5. Spacio-temporal region segmentation for deformable target,6. Development of mobile robot which has 0nboard vision processing system,7. ExperimentsFinally, using the techniques that have developed, mobile robot control under human being walking environment is achieved. Throughout from experiment, evaluation of the techniques and improvement are reported.

实际上，从安全的角度考虑，机器人的工作空间与人类的生活空间是分离的。为了在人类生活中利用机器人，需要检测人类的传感技术。本研究的目的是开发一个机器人视觉系统，可以检测步行或移动的人，而机器人也在移动，通过结合颜色分割，光流生成和深度图生成。以下是研究主题列表，1.颜色分割方法从相关性计算，2.基于光流场产生的边界检测; 3.从立体相机生成深度图，4.高速三轴机器人机头的开发，5.变形目标的时空区域分割，6.研制了具有机载视觉处理系统的移动的机器人，7.最后，利用所开发的技术，实现了人步行环境下的移动的机器人控制。通过实验，对这些技术进行了评价和改进。

项目成果

H. Mizoguchi, Y. Kiuchi, T. Shigehara, K. Hiraoka, and T. Mishima: "Simulation of Microphones Array to Form Acoustic Focus"Proceedings of the IASTED International Conference on Applied Modeling and Simulation (AMS'99). 412-416 (1999)

H. Mizoguchi、Y. Kiuchi、T. Shigehara、K. Hiraoka 和 T. Mishima：“模拟麦克风阵列形成声学焦点”IASTED 国际应用建模与仿真会议 (AMS99) 会议记录。

Kagami,Okada,Inoue,et al.: "Real-Time 3D Optical Flow Generetion System"Proc.of International Conference on Multisensor Fusion and Integration for Intellgent Systems (MFI'99). 1. 237-242 (1999)

Kagami，Okada，Inoue，等人：“实时3D光流生成系统”智能系统多传感器融合与集成国际会议（MFI99）的Proc.。

Satoshi KAGAMI, Kei OKADA, Mitsutaka KABASAWA, Yoshio MATSUMOTO, Atsushi KONNO, Masayuki INABA, Hirochika INOUE: "A Vision-based Legged Robot as a Research Platform"Proc. of International Conference on Intelligent Robots and Systems (IROS'98). vol. 1. 235

Satoshi KAGAMI、Kei OKADA、Mitsutaka KABASAWA、Yoshio MATSUMOTO、Atsushi KONNO、Masayuki Inaba、Hirochika INOUE：“基于视觉的腿式机器人作为研究平台”Proc。

Y. Goto, K. Hidai, H. Mizoguchi, T. Shigehara, and T. Mishima: "Vision Module to Estimate Robot's Self Moving Variation"Proceedings of the 13th Korean Automatic Control Conference ('98KACC). 89-94 (1998)

Y. Goto、K. Hidai、H. Mizoguchi、T. Shigehara 和 T. Mishima：“估计机器人自移动变化的视觉模块”第 13 届韩国自动控制会议 (98KACC) 会议记录。

S. Akaho, Y. Goto, T. Kurita, and H. Mizoguchi: "Pursuit Movement of Pan-Tilt Camera by Feedback-Error-Learning"Proceedings of the Fifth International Conference on Neural Information Processing (ICONIP'98). 409-412 (1998)

S. Akaho、Y. Goto、T. Kurita 和 H. Mizoguchi：“通过反馈误差学习追踪云台摄像机”第五届国际神经信息处理会议 (ICONIP98) 论文集。