• The Journal of Korea Robotics Society
• /
• v.17 no.3
• /
• pp.255-263
• /
• 2022

3D object detection pipelines often incorporate RGB-based object detection methods such as YOLO, which detects the object classes and bounding boxes from the RGB image. However, in complex environments where objects are heavily cluttered, bounding box approaches may show degraded performance due to the overlapping bounding boxes. Mask based methods such as Mask R-CNN can handle such situation better thanks to their detailed object masks, but they require much longer time for data preparation compared to bounding box-based approaches. In this paper, we present a 3D object recognition pipeline which uses either the YOLO or Mask R-CNN real-time object detection algorithm, K-nearest clustering algorithm, mask reduction algorithm and finally Principal Component Analysis (PCA) alg orithm to efficiently detect 3D poses of objects in a complex environment. Furthermore, we also present an improved YOLO based 3D object detection algorithm that uses a prioritized heightmap clustering algorithm to handle overlapping bounding boxes. The suggested algorithms have successfully been used at the Artificial-Intelligence Robot Challenge (ARC) 2021 competition with excellent results.

• The Journal of Korea Robotics Society
• /
• v.4 no.4
• /
• pp.298-304
• /
• 2009

This paper describes an algorithm that improves 3D reconstruction result using a multi-sensor fusion disparity map. We can project LRF (Laser Range Finder) 3D points onto image pixel coordinatesusing extrinsic calibration matrixes of a camera-LRF (${\Phi}$, ${\Delta}$) and a camera calibration matrix (K). The LRF disparity map can be generated by interpolating projected LRF points. In the stereo reconstruction, we can compensate invalid points caused by repeated pattern and textureless region using the LRF disparity map. The result disparity map of compensation process is the multi-sensor fusion disparity map. We can refine the multi-sensor 3D reconstruction based on stereo vision and LRF using the multi-sensor fusion disparity map. The refinement algorithm of multi-sensor based 3D reconstruction is specified in four subsections dealing with virtual LRF stereo image generation, LRF disparity map generation, multi-sensor fusion disparity map generation, and 3D reconstruction process. It has been tested by synchronized stereo image pair and LRF 3D scan data.

• The Journal of Korea Robotics Society
• /
• v.18 no.3
• /
• pp.337-345
• /
• 2023

In this paper, we propose a 3D LiDAR sensor-based costmap generation and path planning algorithm using it for reliable autonomous flight in complex indoor environments. 3D path planning is essential for reliable operation of UAVs. However, existing grid search-based or random sampling-based path planning algorithms in 3D space require a large amount of computation, and UAVs with weight constraints require reliable path planning results in real time. To solve this problem, we propose a method that divides a 3D space into several 2D spaces and a path planning algorithm that considers the distance to obstacles within each space. Among the paths generated in each space, the final path (Best path) that the UAV will follow is determined through the proposed objective function, and for this purpose, we consider the rotation angle of the 2D space, the path length, and the previous best path information. The proposed methods have been verified through autonomous flight of UAVs in real environments, and shows reliable obstacle avoidance performance in various complex environments.

• 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.

• Korean Journal of Computational Design and Engineering
• /
• v.21 no.1
• /
• pp.78-89
• /
• 2016

When a plant owner requests plant 3D CAD models in the format that a shipbuilding company does not use, the shipyard manually re-models plant 3D CAD models according to the owner's requirement. Therefore, it is important to develop a technology to compare the re-modeled plant 3D CAD models with original ones and to quantitatively evaluate similarity between two models. In the previous study, we developed a graphic user interface (GUI)-based comparison system where a user evaluates similarity between original and re-modeled plant 3D CAD models for piping design at the level of unit. However, an offshore plant consists of thousands of units and thus a system which compares several plant 3D CAD models at unit-level without human intervention is necessary. For this, we developed a new batch model comparison system which automatically evaluates similarity of several unit-level plant 3D CAD models using an extensible markup language (XML) file storing file location and name data about a set of plant 3D CAD models. This paper suggests system configuration of a batch-mode-based comparison system and discusses its core functions. For the verification of the developed system, comparison experiments for offshore plant 3D piping CAD models using the system were performed. From the experiments, we confirmed that similarities for several plant 3D CAD models at unit-level were evaluated without human intervention.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.40 no.3
• /
• pp.602-612
• /
• 2015

Current IEEE 802.11p which is suggested for vehicle to vehicle communication supports one hop communication. Thus, it has a limitation to carry out efficient data dissemination. In this thesis, we suggest LTE-D2D based vehicle network to provide efficient data dissemination in the vehicle environment. In this network architecture, we use name based message with IP packet options and we put the intermediate vehicle node called 'super vehicle node' and each normal vehicle node in the cell requests data to the super vehicle node. The super vehicle node responses data to the normal vehicle node. Performance analysis is based mathematical modeling. We compare LTE cellular network to LTE-D2D based vehicle network about throughput according to packet delivery time.

• Journal of Korean Society for Geospatial Information Science
• /
• v.17 no.1
• /
• pp.157-166
• /
• 2009

There is a growing interest in ubiquitous-related research and applications. Among them, GPS-based LBS have been developed and used actively. Recently, with the increase of large size buildings and disastrous events, indoor spaces are getting attention and related research activities are being carried out. Core technologies regarding indoor applications may include 3D indoor data modeling and localization sensor techniques that can integrate with indoor data. However, these technologies have not been standardized and established enough to be applied to indoor implementation. Thus, in this paper, we propose a method to build a relatively simple 3D indoor data modeling technique that can be applied to indoor location based applications. The proposed model takes the form of 2D-based multi-layered structure and has capability for 2D and 3D visualization. We tested three prototype applications using the proposed model; CA(cellular automata)-based 3D evacuation simulation, network-based routing, and indoor moving objects tracking using a stereo camera.

• The Journal of the Korea Contents Association
• /
• v.15 no.2
• /
• pp.31-39
• /
• 2015

2D-to-3D video conversion vests 3D effects in a 2D video by generating stereoscopic views using depth cues inherent in the 2D video. This technology would be a good solution to resolve the problem of 3D content shortage during the transition period to the full ripe 3D video era. In this paper, a low-complexity depth generation method for 2D-to-3D video conversion is presented. For temporal consistency in global depth, a pattern-based depth generation method is newly introduced. A low-complexity refinement algorithm for local depth is also provided to improve 3D perception in object regions. Experimental results show that the proposed method outperforms conventional methods in terms of complexity and subjective quality.

• Proceedings of the KSRS Conference
• /
• 2002.10a
• /
• pp.880-884
• /
• 2002

According to the development of computer technology, especially in 3D graphics and visualization, the interest for 3D GIS has been increasing. Several commercial GIS softwares are ready to provide 3D function in their traditional 2D GIS. However, most of these systems are focused on visualization of 3D objects and supports few analysis functions. Therefore in this paper, we design not only a spatial operation processor which can support spatial analysis functions as well as 3D visualization, but also implement a prototype to operate them. In order to support interoperability between the existing models, the proposed spatial operation processor supports the 3D spatial operations based on 3D geometry object model which is designed to extend 2D geometry model of OGIS consortium, and supports index based on R$^*$-Tree. The proposed spatial operation processor can be applied in 3D GIS to support 3D analysis functions.

• Journal of Fashion Business
• /
• v.22 no.4
• /
• pp.93-105
• /
• 2018

Using FDM 3D printing, yarn shape and composition were modeled and 3D printed with PLA and TPU filaments currently used for apparel. Based on this, mechanical characteristics were measured to determine 3D printing yarn according to type of filaments in the 3D printed output and deformation and recovery characteristics due to differences in structure type. As a result of examining tensile and shear characteristics of PLA and TPU 3D printing compiles, TPU overall was measured with significantly lower stress than PLA. This is due to high elasticity of TPU's character, revealing that it has better flexibility than PLA. In addition, during deformation due to external forces, the more freedom between the head and foot parts of the loop, and the lower the force associated with each other, the more flexible it is. TPU revealed that it was easier to tension and recovery from tensile deformation than PLA, indicating potential for clothing materials using 3D printing. If high-molecular materials, such as PLA flexibility, it is likely to provide some flexibility through development of styles, including degree of freedom in modeling. Based on this, we provide basic data for developing 3D printing textures that can be satisfied with textile for apparel.