Bum-Jae You

Three-dimensional pose determination for a humanoid robot using binocular head system

There has been much interests on three-dimensional pose estimation of an object since there are a lot of robotic applications such as intelligent robotic assembly and/or vision-guided control of a humanoid or human-friendly robot. In this paper, there are proposed a simple and fast stereo matching algorithm for real-time robotic applications and an improved scheme for feature matching using three-dimensional information by adopting backprojection as a feedback element measuring the degree of feature matching. The degree of feature matching is determined by checking existence ratio of each feature point in real images after back-projecting an object model into image plane. The proposed pose determination algorithm is applied for a humanoid robot in KIST, whose name is CENTAUR, successfully and determine the pose of several polyhedral objects using 3D information of vertexes on the outline of an object in image plane.

Development of an Unmanned Autonomous Concrete Floor Robotic Troweling System

It is natural to introduce an automatic floor troweling system to a spacious concrete floor from the point of construction cost and time. Among several methods it would be the efficient to modify an existing available ride-on trowel into an unmanned and remote controlled troweling robot. The handling mechanism of a power trowel is converted to electric actuation of DC servo motors. Overall control system is designed to have a network-based real- time distributed architecture under CAN in order to distract the computational load concentrated on central main controller. The remote controller not only contains many functions to direct the troweling robot in the same way as riding operator does, but also works as input and monitoring devices in an autonomous mode. The positioning system of the laser range scanning sensor and several reflective beacons, and the job planner based on a blueprint of the floor provides the troweling robot with intelligence to work and navigate autonomously within a given floor area.

Butler: A Visitor Guide Robot

This paper describes the major components of software architecture of an autonomous visitor guide robot with the brief explanation of its hardware system. This robot, “Butler,” was deployed in a real office environment, Intelligent Robot Research Center at KIST, where it guided many visitors successfully. To make the robot navigate the office and interact with visitors without problems, many challenges such as navigation in dynamic environments and interaction through various methods should be overcome. In this paper, the solutions for those problems are illustrated and our conclusion about this project is discussed.

High-Speed Visual Measurement System of Pile Penetration and Rebound Movement for Construction

A number of piles should be driven into the ground by a hammering process in order to make the ground under the structure safe and strong when construction companies build a high structure such as building and bridges. It is essential to determine whether piles are penetrated into the ground enough to support the weight of the structure since ground characteristics at different locations are different each other. This paper proposes a high-speed real-time visual measurement approach for pile movement under hammering by combining a high-speed line-scan camera with a specially designed mark to recognize two-dimensional motion parameters, position and orientation, of a pile. A mark stacking white and black right-angled triangles is used for the measurement, and movement information for vertical distance, horizontal distance and rotational angle is determined simultaneously. Especially, a high-speed line-scan CCD camera whose line rate is greater than 10 KHz improves the measurement performance of dynamic characteristics of a pile at impact instant dramatically.

Fast Markov Localization in Indoor Environment Using SFM Angle-Histogram

Localization is one of the most important issues for mobile robots since all tasks are commanded to a mobile robot based on the assumption that the mobile robot knows its position. Even though non-probabilistic techniques are faster than probabilistic approaches, those are sensitive to measurement errors and a mobile robot may lose its position in complex environments. And most simple features need additional information to represent the characteristics of environments. On the contrary, probabilistic approaches have many advantages since those can cope with sensor noises and can globally localize a mobile robot. However, those probabilistic approaches are time-consuming techniques because of the heavy computational loads due to huge comparative data. In this paper, we propose a fast probabilistic localization method including global localization by remodeling raw laser sensory data using angle histogram to reduce computational loads for localization. The algorithm is experimented successfully by using a mobile robot named KARA. Keywordsglobal localization, fast probabilistic localization, sensor remodeling, angle-histgram