• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.11 no.7
• /
• pp.2579-2585
• /
• 2010

This paper proposes a dynamic human pose tracking method using motion-based search strategy from an image sequence obtained from a monocular camera. The proposed method compares the image features between 3D human model projections and real input images. The method repeats the process until predefined criteria and then estimates 3D human pose that generates the best match. When searching for the best matching configuration with respect to the input image, the search region is determined from the estimated 2D image motion and then search is performed randomly for the body configuration conducted within that search region. As the 2D image motion is highly constrained, this significantly reduces the dimensionality of the feasible space. This strategy have two advantages: the motion estimation leads to an efficient allocation of the search space, and the pose estimation method is adaptive to various kinds of motion.

• Journal of Institute of Control, Robotics and Systems
• /
• v.20 no.2
• /
• pp.218-222
• /
• 2014

In this paper, we introduce an autonomous calibration method for a 2D laser displacement sensor (e.g. laser vision sensor and laser range finder) by matching a single point on a flat structure. Many arc welding robots install a 2D laser displacement sensor to expand their application by recognizing their environment (e.g. base metal and seam). In such systems, sensing data should be transformed to the robot's coordinates, and the geometric relation (i.e. rotation and translation) between the robot's coordinates and sensor coordinates should be known for the transformation. Calibration means the inference process of geometric relation between the sensor and robot. Generally, the matching of more than 3 points is required to infer the geometric relation. However, we introduce a novel method to calibrate using only 1 point matching and use a specific flat structure (i.e. circular hole) which enables us to find the geometric relation with a single point matching. We make the rotation component of the calibration results as a constant to use only a single point by moving a robot to a specific pose. The flat structure can be installed easily in a manufacturing site, because the structure does not have a volume (i.e. almost 2D structure). The calibration process is fully autonomous and does not need any manual operation. A robot which installed the sensor moves to the specific pose by sensing features of the circular hole such as length of chord and center position of the chord. We show the precision of the proposed method by performing repetitive experiments in various situations. Furthermore, we applied the result of the proposed method to sensor based seam tracking with a robot, and report the difference of the robot's TCP (Tool Center Point) trajectory. This experiment shows that the proposed method ensures precision.

• Journal of KIISE:Software and Applications
• /
• v.30 no.9
• /
• pp.829-842
• /
• 2003

In this Paper, we propose the new face detection and tracking method based on template matching for real-time applications such as, teleconference, telecommunication, front stage of surveillance system using face recognition, and video-phone applications. Since the main purpose of paper is to track a face regardless of various environments, we use template-based face tracking method. To generate robust face templates, we apply wavelet transform to the average face image and extract three types of wavelet template from transformed low-resolution average face. However template matching is generally sensitive to the change of illumination conditions, we apply Min-max normalization with histogram equalization according to the variation of intensity. Tracking method is also applied to reduce the computation time and predict precise face candidate region. Finally, facial components are also detected and from the relative distance of two eyes, we estimate the size of facial ellipse.

• Journal of Institute of Control, Robotics and Systems
• /
• v.19 no.8
• /
• pp.745-754
• /
• 2013

An efficient urban outdoor localization and navigation system is proposed for car-like robots. First an accurate outdoor localization method is suggested using line/arc features and 2.5D map matching with LRFs (Laser Range Finders), which can reduce the number of singular cases and increase accuracy. Also, path generation, path tracking, and path modification algorithms are proposed for navigation. All these algorithms are implemented on an electric scooter to construct an autonomous urban outdoor localization and navigation system. Experiments reveal the practicality of the proposed system.

• Journal of Information Processing Systems
• /
• v.15 no.5
• /
• pp.1218-1230
• /
• 2019

This study proposes a novel video traffic flow detection method based on machine vision technology. The three-frame difference method, which is one kind of a motion evaluation method, is used to establish initial background image, and then a statistical scoring strategy is chosen to update background image in real time. Finally, the background difference method is used for detecting the moving objects. Meanwhile, a simple but effective shadow elimination method is introduced to improve the accuracy of the detection for moving objects. Furthermore, the study also proposes a vehicle matching and tracking strategy by combining characteristics, such as vehicle's location information, color information and fractal dimension information. Experimental results show that this detection method could quickly and effectively detect various traffic flow parameters, laying a solid foundation for enhancing the degree of automation for traffic management.

• The KIPS Transactions:PartB
• /
• v.11B no.1
• /
• pp.27-34
• /
• 2004

This paper proposes a method to trace and interpret a moving object based on the information which can be directly obtained from MPEG-2 compressed video stream without decoding process. In the proposed method, the motion flow is constructed from the motion vectors included in compressed video. We calculate the amount of pan, tilt, and zoom associated with camera operations using generalized Hough transform. The local object motion can be extracted from the motion flow after the compensation with the parameters related to the global camera motion. Initially, a moving object to be traced is designated by user via bounding box. After then automatic tracking Is performed based on the accumulated motion flows according to the area contributions. Also, in order to reduce the cumulative tracking error, the object area is reshaped in the first I-frame of a GOP by matching the DCT coefficients. The proposed method can improve the computation speed because the information can be directly obtained from the MPEG-2 compressed video, but the object boundary is limited by macro-blocks rather than pixels. Also, the proposed method is proper for approximate object tracking rather than accurate tracing of an object because of limited information available in the compressed video data.

• Proceedings of the IEEK Conference
• /
• 2002.06c
• /
• pp.13-16
• /
• 2002

This paper proposes a method to track hand gesture and to recognize the gesture pattern using Kalman filter and continuous dynamic programming (CDP). The positions of hands are predicted by Kalman filter, and corresponding pixels to the hands are extracted by skin color filter. The center of gravity of the hands is the same as the input pattern vector. The input gesture is then recognized by matching with the reference gesture patterns using CDP. From experimental results to recognize circle shape gesture and intention gestures such as “Come on” and “Bye-bye”, we show the proposed method is feasible to the hand gesture-based human -computer interaction.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 1996.11a
• /
• pp.419-424
• /
• 1996

As track density increases, the effects of nonlinearity in pivot bearing of hard disk drive on the servo performance are becoming more important in considering the range of inertia force and the input torque during settling and tracking mode. Recently, an increasing attention is given to more precise experimental observations and modelings of pivot nonlinearity for achieving higher performance of servo control. In this paper, we propose a new model that shows an improved prediction of the pivot nonlinearity than existing preload-plus-two-slope model at matching simulations and experimental results in both time and frequency domains. Experimental measurements are carried out to validate and identify the specific nonlinearity presents in the pivot bearing when its in fine motion. Using the experimental results new model along with the existing one are characterized and compared for relevancies.

• Journal of Institute of Control, Robotics and Systems
• /
• v.15 no.5
• /
• pp.473-479
• /
• 2009

This paper proposes an efficient method to locate the automated guided vehicle (AGV) into a specific parking position using artificial visual landmark and vision-based algorithm. The landmark has comer features and a HSI color arrangement for robustness against illuminant variation. The landmark is attached to left of a parking spot under a crane. For parking, an AGV detects the landmark with CCD camera fixed to the AGV using Harris comer detector and matching descriptors of the comer features. After detecting the landmark, the AGV tracks the landmark using pyramidal Lucas-Kanade feature tracker and a refinement process. Then, the AGV decreases its speed and aligns its longitudinal position with the center of the landmark. The experiments showed the AGV parked accurately at the parking spot with small standard deviation of error under bright illumination and dark illumination.

• Journal of Institute of Control, Robotics and Systems
• /
• v.17 no.9
• /
• pp.902-907
• /
• 2011

In this paper, we deal with performance evaluation of visual path following using 2D and 3D information. Visual path follow first teaches driving path by selecting milestone images then follows the same route by comparing the milestone image and current image. We follow the visual path following algorithm of [8] and [10]. In [8], a robot navigated with 2D image information only. But in [10], local 3D geometries are reconstructed between the milestone images in order to achieve fast feature prediction which allows the recovery from tracking failures. Experimental results including diverse indoor cases show performance of each algorithm.