• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2009.01a
• /
• pp.808-811
• /
• 2009

Robovie-R2 [1], developed by ATR, is a 110cm high, 60kg weight, two wheel drive, human like robot. It has two arms with dynamic fingers. It also has a position sensitive detector sensor and two cameras as eyes on his head for recognizing his surrounding environment. Recent years, we have carried out a project to integrate new functions into Robovie-R2 so as to make it possible to be used in a dining room in healthcare center for helping serving meal for elderly. As a new function, we have developed software system for adaptive movement control of Robovie-R2 that is primary important since a robot that cannot autonomously control its movement would be a dangerous object to the people in dining room. We used the cameras on Robovie-R2's head to catch environment images, applied our original algorithm for recognizing obstacles such as furniture or people, so as to control Roboie-R2's movement. In this paper, we will focus our algorithm and its results.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.23 no.6
• /
• pp.527-532
• /
• 2013

Depth estimation is an essential factor for robot vision, 3D scene modeling, and motion control. The depth estimation method is based on focusing values calculated in a series of images by a single camera at different distance between lens and object. In this paper, we proposed a relative depth estimation method using focus measure. The proposed method is implemented by focus value calculated for each image obtained at different lens position and then depth is finally estimated by considering relative distance of two patterns. We performed various experiments on the effective focus measures for depth estimation by using various patterns and their usefulness.

• Journal of Internet Computing and Services
• /
• v.12 no.5
• /
• pp.101-111
• /
• 2011

Most researches in WSN-based parking management system used wireless sensors to monitor the events in a car parking area. However, the problem of car collisions in car parks was not discussed by previous researches. The car position details over time are vital in analyzing a collision event. This paper proposes a collision verification method to detect and to analyze the collision event in the parking area, and then notifies car owners. The detection uses the information from motion sensors for comprehensive details of position and direction of a moving car, and the verification processes an object tracking technique with a fast OBB intersection test. The performance tests show that the location technique is more accurate with additional sensors and the OBB collision test is faster compared to a normal OBB intersection test.

• Journal of the Korea Society of Computer and Information
• /
• v.12 no.2 s.46
• /
• pp.131-139
• /
• 2007

This paper proposed a multi-resolution algorithm. Its search point and complexity were compared with those of block match algorithm. Also the speed up comparison was made with the block match algorithm. The proposed multi-resolution NTSS-3 Level algorithm was compared again with its targets, TSS-3 Level algorithm and NTSS algorithm. The comparison results showed that the NTSS-3 Level algorithm was superior in search point and speed up. Accordingly, the proposed NTSS-3 Level algorithm was two to three times better in search point and two to four times better in complexity calculation than those of the compared object, the block match algorithm. In speed up, the proposed NTSS-3 Level algorithm was two times better. Accordingly, the proposed multi-resolution NTSS-3 Level algorithm showed PSNR ration portion excellency in search point and speed up.

• Annals of Clinical Neurophysiology
• /
• v.1 no.2
• /
• pp.173-181
• /
• 1999

Eye movements serve vision by placing the image of an object on the fovea of each retina, and by preventing slippage of images on the retina. The brain employs two modes of ocular motor control, fast eye movements (saccades) and smooth eye movements. Saccades bring the fovea to a target, and smooth eye movements prevent retinal image slip. Smooth eye movements comprise smooth pursuit, the optokinetic reflex, the vestibulo-ocular reflex (VOR), vergence, and fixation. Saccades achieve rapid refixation of targets that fall on the extrafoveal retina by moving the eyes at peak velocities that can exceed $700^{\circ}/s$. Various brain lesions can affect saccadic latency, velocity, or accuracy. Smooth pursuit maintains fixation of a slowly moving target. The pursuit system responds to slippage of an image near the fovea in order to accelerate the eyes to a velocity that matches that of the target. When smooth eye movements velocity fails to match target velocity, catch-up saccades are used to compensate for limited smooth pursuit velocities. The VOR subserves vision by generating conjugate eye movements that are equal and opposite to head movements. If the VOR gain (the ratio of eye velocity to head velocity) is too high or too low, the target image is off the fovea, and head motion causes oscillopsia, an illusory to-and-fro movement of the environment.

• Journal of the Korean Society for Precision Engineering
• /
• v.15 no.6
• /
• pp.107-115
• /
• 1998

A path planning using potential field method is very useful for the real-time navigation of mobile robots. However, the method needs high modeling cost to calculate the potential field because of complex preprocessing, and mobile robots may get stuck into local minima. In this paper, An efficient path planning algorithm for multiple mobile robots, based on the potential field method, was proposed. In the algorithm. the concepts of subgoals and obstacle priority were introduced. The subgoals can be used to escape local minima, or to design and change the paths of mobile robots in the work space. In obstacle priority, all the objects (obstacles and mobile robots) in the work space have their own priorities, and the object having lower priority should avoid the objects having higher priority than it has. In this paper, first, potential based path planning method was introduced, next an efficient collision-avoidance algorithm for multiple mobile robots, moving in the obstacle environment, was proposed by using subgoals and obstacle priority. Finally, the developed algorithm was demonstrated graphically to show the usefulness of the algorithm.

• The Journal of Korea Robotics Society
• /
• v.10 no.4
• /
• pp.200-212
• /
• 2015

In classical dynamics, the coefficient of restitution is one of variables to estimate the amount of impulse. In general, we have considered the coefficient of restitution as a constant value. However, coefficient of restitution (COR) is the function of contact material and colliding velocity. Furthermore, COR is also a function of contact area. Thus, without considering the variable characteristic of COR, the actual motion of an object just after impact is not the same as we expect. A general COR model is proposed in this work and its effectiveness is verified through a cart impact experiment and its result is applied to simulation of a ball impact problem. A three-degree-of-freedom manipulator is employed as a test-bed.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.10 no.11
• /
• pp.5624-5638
• /
• 2016

A novel method based on Surf points is proposed to detect and lock-track single ground target in aerial videos. Videos captured by moving cameras contain complex motions, which bring difficulty in moving object detection. Our approach contains three parts: moving target template detection, search area estimation and target tracking. Global motion estimation and compensation are first made by grids-sampling Surf points selecting and matching. And then, the single ground target is detected by joint spatial-temporal information processing. The temporal process is made by calculating difference between compensated reference and current image and the spatial process is implementing morphological operations and adaptive binarization. The second part improves KALMAN filter with surf points scale information to predict target position and search area adaptively. Lastly, the local Surf points of target template are matched in this search region to realize target tracking. The long-term tracking is updated following target scaling, occlusion and large deformation. Experimental results show that the algorithm can correctly detect small moving target in dynamic scenes with complex motions. It is robust to vehicle dithering and target scale changing, rotation, especially partial occlusion or temporal complete occlusion. Comparing with traditional algorithms, our method enables real time operation, processing $520{\times}390$ frames at around 15fps.

• Journal of Korean Society for Geospatial Information Science
• /
• v.1 no.2 s.2
• /
• pp.153-158
• /
• 1993

It is important to establish the transformational relation between scanned image coordinates and digital image coordinates because the coordinate system of digital image is transformed from scanned image coordinate system through scanning work. And, some researches are required in scanning works to correct the deformation that is due to the motion of scanner. In this study, some procedures are applied to determine the optimal calibration model equation which can calibrate the scanner. As a result the optimal calibration model equation for the object scanner is determined The procedure of this study can applied to the calibration of other types of scanner, because the procedures are done with the analysis of geometrical properties rather than the analysis of physical properties.

• Journal of the Korea Society of Computer and Information
• /
• v.13 no.1
• /
• pp.75-80
• /
• 2008

This paper proposes a method to detect scene change using global decision tree that extract boundary cut that have width of big change that happen by camera brake from difference value of frames. First, calculate frame difference value through regional X2-histogram and normalization, next, calculate distance between difference value using normalization. Shot boundary detection is performed by compare global threshold distance with distance value for two adjacent frames that calculating global threshold distance based on distance between calculated difference value. Global decision tree proposed this paper can detect easily sudden scene change such as motion from object or camera and flashlight.