• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.2
• /
• pp.559-575
• /
• 2023

3D human pose estimation is widely applied in various fields, including action recognition, sports analysis, and human-computer interaction. 3D human pose estimation has achieved significant progress with the introduction of convolutional neural network (CNN). Recently, several researches have proposed the use of multiview approaches to avoid occlusions in single-view approaches. However, as the number of cameras increases, a 3D pose estimation system relying on a CNN may lack in computational resources. In addition, when a single host system uses multiple cameras, the data transition speed becomes inadequate owing to bandwidth limitations. To address this problem, we propose a distributed real-time 3D pose estimation framework based on asynchronous multiple cameras. The proposed framework comprises a central server and multiple edge devices. Each multiple-edge device estimates a 2D human pose from its view and sendsit to the central server. Subsequently, the central server synchronizes the received 2D human pose data based on the timestamps. Finally, the central server reconstructs a 3D human pose using geometrical triangulation. We demonstrate that the proposed framework increases the percentage of detected joints and successfully estimates 3D human poses in real-time.

• Journal of KIBIM
• /
• v.5 no.2
• /
• pp.46-54
• /
• 2015

Vision-based tracking has been proposed as a means to efficiently track a large number of construction resources operating in a congested site. In order to obtain 3D coordinates of an object, it is necessary to employ stereo-vision theories. Detecting and tracking of multiple objects require an entity matching process that finds corresponding pairs of detected entities across the two camera views. This paper proposes an efficient way of entity matching for tracking of construction workers. The proposed method basically uses epipolar geometry which represents the relationship between the two fixed cameras. Each pixel coordinate in a camera view is projected onto the other camera view as an epipolar line. The proposed method finds the matching pair of a worker entity by comparing the proximity of the all detected entities in the other view to the epipolar line. Experimental results demonstrate its suitability for automated entity matching for 3D vision-based tracking of construction workers.

• Journal of Broadcast Engineering
• /
• v.13 no.6
• /
• pp.859-874
• /
• 2008

In this paper, we develop a fast view synthesis method that generates multiple intermediate views in real-time for the 3D display system when the camera geometry and depth map of reference views are given in advance. The proposed method achieves faster view synthesis than previous approaches in GPU by processing in parallel the entire computations required for the view synthesis. Specifically, we use $CUDA^{TM}$ (by NVIDIA) to control GPU device. For increasing the processing speed, we adapted all the processes for the view synthesis to single instruction multiple data (SIMD) structure that is a main feature of CUDA, maximized the use of the high-speed memories on GPU device, and optimized the implementation. As a result, we could synthesize 9 intermediate view images with the size of 720 by 480 pixels within 0.128 second.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.45 no.6
• /
• pp.44-56
• /
• 2008

Previous hardware devices to capture human motion have many limitations; expensive equipment, complexity of manipulation or constraints of human motion. In order to overcome these problems, real-time motion capture algorithms based on computer vision have been actively proposed. This paper presents an efficient analysis method of multiple view images for real-time motion capture. First, we detect the skin color regions of human being, and then correct the image coordinates of the regions by using camera calibration and epipolar geometry. Finally, we track the human body part and capture human motion using kalman filter. Experimental results show that the proposed algorithm can estimate a precise position of the human body.

• Journal of Institute of Control, Robotics and Systems
• /
• v.14 no.7
• /
• pp.635-641
• /
• 2008

In this paper, we used a laser range finder (LRF) to detect both the static and dynamic obstacles for the safe navigation of a mobile robot. LRF sensor measurements containing the information of obstacle's geometry are first processed to extract the characteristic points of the obstacle in the sensor field of view. Then the dynamic states of the characteristic points are approximated using kinematic model, which are tracked by associating the measurements with Probability Data Association Filter. Finally, the collision avoidance algorithm is developed by using fuzzy decision making algorithm depending on the states of the obstacles tracked by the proposed obstacle tracking algorithm. The performance of the proposed algorithm is evaluated through experiments with the experimental mobile robot.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.49 no.4
• /
• pp.10-16
• /
• 2012

For high speed visual inspection in semiconductor industries, it is essential to acquire two-dimensional images on regions of interests with a large field of view (FOV) and a high resolution simultaneously. In this paper, an imaging system is newly proposed to achieve high quality image in terms of precision and FOV, which is composed of single lens, a beam splitter, two camera sensors, and stereo image grabbing board. For simultaneously acquired object images from two camera sensors, Zhang's camera calibration method is applied to calibrate each camera first of all. Secondly, to find a mathematical mapping function between two images acquired from different view cameras, the matching matrix from multiview camera geometry is calculated based on their image homography. Through the image homography, two images are finally registered to secure a large inspection FOV. Here the inspection system of using multiple images from multiple cameras need very fast processing unit for real-time image matching. For this purpose, parallel processing hardware and software are utilized, such as Compute Unified Device Architecture (CUDA). As a result, we can obtain a matched image from two separated images in real-time. Finally, the acquired homography is evaluated in term of accuracy through a series of experiments, and the obtained results shows the effectiveness of the proposed system and method.

• Proceedings of the KWS Conference
• /
• 2002.10a
• /
• pp.752-756
• /
• 2002

Variable weld bead penetrations related to the base metal chemistry of stainless steels in GTA welding have been under constant investigations due to their industrial implications. It has been proposed that among other elements, the sulfur content of steels determines the weld pool geometry, particularly its penetration. It is suggested that the surface tension temperature gradient of steels becomes positive with appropriate dosing in sulfur and results in inward melt flow, propitious for deeper welds. However, the chemistry of industrial steels is complex due to the presence of multiple minor elements either deliberately added or remnant impurity traces. With this in view, investigations on 41 austenitic and nine martensitic stainless steels were carried to see if there existed any possible relation between the weld profile and some of the designated elements. The results suggest no direct correlation between sulfur or any other major or trace element and weld penetration. At first glance the results are contradictory to what is often asserted.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2006.05a
• /
• pp.1387-1390
• /
• 2006

Several actuation mechanism for carton nanotubes has teen reported recently, including actuation by double-layer charge injection and ac voltages applyied to multiple electrodes. Carbon nanotube actuator based on double layer charge injection work well in electrolyte at low voltage. AC dielectrophoresis based on four electrode geometry demonstrated carton nanotubes in solution phase can be oriently manipulated by dielectrophoresis. From this point of view, and in regard to their performance, bucky paper actuator may alternate natural muscle. also, applied AC signal with appropriate magnitude and frequency together with four electrode arrangement has potential to realize nanotube electrokinetics.

• Journal of Platform Technology
• /
• v.6 no.4
• /
• pp.18-25
• /
• 2018

This paper describes a real-time human and robot tracking method in Intelligent Space with multi-camera networks. The proposed method detects candidates for humans and robots by using the histogram of oriented gradients (HOG) feature in an image. To classify humans and robots from the candidates in real time, we apply cascaded structure to constructing a strong classifier which consists of many weak classifiers as follows: a linear support vector machine (SVM) and a radial-basis function (RBF) SVM. By using the multiple view geometry, the method estimates the 3D position of humans and robots from their 2D coordinates on image coordinate system, and tracks their positions by using stochastic approach. To test the performance of the method, humans and robots are asked to move according to given rectangular and circular paths. Experimental results show that the proposed method is able to reduce the localization error and be good for a practical application of human-centered services in the Intelligent Space.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.47 no.2
• /
• pp.20-32
• /
• 2010

Previous hardware devices to capture human motion have many limitations; expensive equipment, complexity of manipulation or constraints of human motion. In order to overcome these problems, real-time motion capture algorithms based on computer vision have been actively proposed. This paper presents an efficient analysis method of multiple view images for real-time motion capture. First, we detect the skin color regions of human being, and then correct the image coordinates of the regions by using camera calibration and epipolar geometry. Finally, we track the human body part and capture human motion using kalman filter. Experimental results show that the proposed algorithm can estimate a precise position of the human body.