Jin-Oh Kim

Robotic smart house to assist people with movement disabilities

This paper introduces a new robotic smart house, Intelligent Sweet Home, developed at KAIST in Korea, which is based on several robotic agents and aims at testing advanced concepts for independent living of the elderly and people with disabilities. The work focuses on technical solutions for human-friendly assistance in motion/mobility and advanced human-machine interfaces that provide simple control of all assistive robotic systems and home-installed appliances. The smart house concept includes an intelligent bed, intelligent wheelchair, and robotic hoist for effortless transfer of the user between bed and wheelchair. The design solutions comply with most of the users’ requirements and suggestions collected by a special questionnaire survey of people with disabilities. The smart house responds to the users commands as well as to the recognized intentions of the user. Various interfaces, based on hand gestures, voice, body movement, and posture, have been studied and tested. The paper describes the overall system structure and explains the design and functionality of some main system components.

Automatic focus control for assembly alignment in a lens module process

This study proposes an auto focusing method for multi-focus images in assembling lens modules in digital camera phones. A camera module in a camera phone is composed of a CCD sensor, an IR glass, a lens and a PCB board. The alignment of the components is one of the most important factors in determining the product quality. Auto-focus is essential to adjust the image quality of the IR glass, but there are two focal points in the captured image, due to the thickness of the IR glass. Therefore, the sharpness, probability and scale factor are used to find the desired focus from a multi-focus image. The sharpness is defined as the clarity of the image. The variation of the probability and scale factors are analyzed to find the focus after pattern recognition with a registered image. The presented algorithm was applied to a lens assembly machine which has 5 axes, two vacuum chucks and an inspection system. The desired focus can be determined on the local maximum of the sharpness, the probability and the scale factor in the experiment.

An attitude control and stabilization of an unstructured object using balancing beam, new construction machinery

In this study, the CMG (control moment gyro) subsystem, a new type of construction machinery using the mechanism of CMG, for the attitude control of an unstructured object such as a beam carried by a tower crane, is designed and implemented. The CMG subsystem consists of one motor for spinning the wheel, and the other motor for turning the outer gimbal. As an input of the system, applying force to the spin axis of the wheel leads the torque about the axis, whereas the wheel is spinning at high speed. This torque is used to control the attitude of the unstructured object in this study. For the stabilizer function, in addition, holding the load at the current position, the power applied to the gimbal motor of CMG will be cut, which results in a braking force to stop the load by the gyro effect. The CMG subsystem the authors developed consists of the mechanical system, a remote controller, a servo controller and control software for the system. The authors also present experimental results to show that the system they proposed is useful as construction machinery which can control the attitude of the beam hanging from a tower crane.

Visual recognition of planar panels with line features

The authors introduce a new method for visual recognition of panels with stiffeners for the sub-assembly process of ship-building. Recognition with existing CAD data is to find the location/orientation of stiffeners and panels, panel ID numbers and approximate welding lines in sequence. The method extracts line features from original images and uses five-line invariants for the recognition. The existing invariant approaches have been proved successful for planar object recognition, but the large search space with combinatorial explosion has been the main hurdle in real application. To reduce this search space, the authors propose a new method with line convex hull (LCH) and indexing logic filter (ILF). In addition, a hierarchical database is constructed to retrieve corresponding model features directly. They have performed experiments using fifteen different welding panels, and demonstrate the feasibility of their method for the recognition of planar objects.