運用紅外線及雷射光於視覺式智慧型空間之機器人定位與地圖重建

本論文提出一套視覺式智慧型空間與移動式機器人相互配合之地圖建立系統，視覺式智慧型空間採用多台IP(Internet protocol)攝影機為感測元件，分別監控著不同但相互有重疊部分的區域且未知彼此位置關係，移動式機器人上安裝兩套視覺系統，分別為紅外線濾鏡搭配主動式紅外線之單眼視覺系統及主動式雷射光之單眼視覺系統。此地圖建立系統由各自的攝影機擷取投射於環境中之光束，並將所擷取到的光束重建於現實三維空間中，藉此重建環境之輪廓，對應於機器人目前所在區域之攝影機，透過放置於機器人上之格點特徵，即時校準出彼此之間位置關係。移動式機器人透過紅外線及雷射光探索與IP攝影機校準資訊相互結合，即建立出以該IP攝影機為主的環境地圖，並藉由移動式機器人搜尋空間攝影機可視範圍重疊區域，求得IP攝影機之間的相對關係，據此可將建立之地圖整合成智慧型空間地圖，使機器人進行控制任務。本論文於室內空間中架設多台IP攝影機以建構智慧型環境，並以自製移動式機器人驗証所提出之地圖建立系統，實現移動式機器人於視覺式智慧型空間之導航與控制。

關鍵字

智慧型空間；地圖重建；移動式機器人；機器人導航；機器人避障

並列摘要

This paper presents an approach for mobile robot localization and map building in a vision-based intelligent spaceusing infrared, laser lighting and vision sensors. The vision-based intelligent space is composed of a network of uncalibrated ceiling-mounted IP cameras with overlapping areas. Two sensing systems are mounted on the mobile robot. One consists of an infrared projector, an infrared filter and a camera. The other utilizes a laser projector and a camera. Map building is performed by capturing the projection of the beam from infrared or laser projector to reconstruct its three-dimensional coordinates. The local map in a camera's field of view is built by the sensing ystem mounted on the robot and the on-line calibrated IP cameras. Accordingly, the local maps can be integrated into a global map for mobile robot control task in the vision-based intelligent space. The proposed system has been successfully validated by navigating a custom mobile robot equipped with the two sensing systems in an indoor vision-based intelligent environment.

並列關鍵字

Intelligent spaces ； Map building ； Mobile robot ； Robot navigation ； Obstacle avoidance

參考文獻

[32] R. G. Brown and P. Y. C. Hwang. Introduction to Random Signals and Applied Kalman Filter. Wiley, 2nd edition, 1992.

[1] T. Sasaki, D. Brscicand, and H. Hashimoto. Implementation of distributed sensor network for intelligent space. In Proc. of the IEEE International Conference on Mechatronics, pages 1–6, Kumamoto Japan, May 2007.

[3] K. Morioka T.-S. Jin and H. Hashimoto. Distributed sensor network for multi-agent motion tracking in intelligent. In Proc. of the SICE-ICASE International

Joint Conference, pages 3716–3721, Busan, Korea, Oct. 2003.

[4] J. Canou, C. Novales, G. Poisson, and P.Marche. Quick primitives extraction using inertia matrix on measures issue from an ultrasonic network. In Proc. of the IEEE International Conference on Robotics and Automation, pages

國際替代計量

運用紅外線及雷射光於視覺式智慧型空間之機器人定位與地圖重建

未授權

主題瀏覽