• Journal of the Korea Society of Computer and Information
• /
• v.28 no.8
• /
• pp.95-101
• /
• 2023

Colmap is one of the innovative artificial intelligence technologies, highly effective as a tool in 3D reconstruction tasks. Moreover, it excels at constructing intricate 3D models by utilizing images and corresponding metadata. Colmap generates 3D models by merging 2D images, camera position data, depth information, and so on. Through this, it achieves detailed and precise 3D reconstructions, inclusive of objects from the real world. Additionally, Colmap provides rapid processing by leveraging GPUs, allowing for efficient operation even within large data sets. In this paper, we have presented a method of collecting 2D images of traditional Korean towers and reconstructing them into 3D models using Colmap. This study applied this technology in the restoration process of traditional stone towers in South Korea. As a result, we confirmed the potential applicability of Colmap in the field of cultural heritage restoration.

• Journal of the Korea Computer Graphics Society
• /
• v.24 no.3
• /
• pp.61-71
• /
• 2018

When triangular meshes are generated from the point clouds in global space reconstructed through camera pose estimation against captured RGB-D streams, the quality of the resulting meshes improves as more triangles are hired. However, for 3D reconstructed models beyond some size threshold, they become to suffer from the ugly-looking artefacts due to the insufficient precision of RGB-D sensors as well as significant burdens in memory requirement and rendering cost. In this paper, for the generation of 3D models appropriate for real-time applications, we propose an effective technique that extracts high-quality textures for moderate-sized meshes from the captured colors associated with the reconstructed point sets. In particular, we show that via a simple method based on the mapping between the 3D global space resulting from the camera pose estimation and the 2D texture space, textures can be generated effectively for the 3D models reconstructed from captured RGB-D image streams.

• Imaging Science in Dentistry
• /
• v.51 no.1
• /
• pp.41-47
• /
• 2021

Purpose: This study aimed to compare the accuracy of 3-dimensional(3D) printed models derived from multidetector computed tomography (MDCT) and cone-beam computed tomography (CBCT) systems with different fields of view (FOVs). Materials and Methods: Five human dry mandibles were used to assess the accuracy of reconstructions of anatomical landmarks, bone defects, and intra-socket dimensions by 3D printers. The measurements were made on dry mandibles using a digital caliper (gold standard). The mandibles then underwent MDCT imaging. In addition, CBCT images were obtained using Cranex 3D and NewTom 3G scanners with 2 different FOVs. The images were transferred to two 3D printers, and the digital light processing (DLP) and fused deposition modeling (FDM) techniques were used to fabricate the 3D models, respectively. The same measurements were also made on the fabricated prototypes. The values measured on the 3D models were compared with the actual values, and the differences were analyzed using the paired t-test. Results: The landmarks measured on prototypes fabricated using the FDM and DLP techniques based on all 4 imaging systems showed differences from the gold standard. No significant differences were noted between the FDM and DLP techniques. Conclusion: The 3D printers were reliable systems for maxillofacial reconstruction. In this study, scanners with smaller voxels had the highest precision, and the DLP printer showed higher accuracy in reconstructing the maxillofacial landmarks. It seemed that 3D reconstructions of the anterior region were overestimated, while the reconstructions of intra-socket dimensions and implant holes were slightly underestimated.

• Korean Journal of Computational Design and Engineering
• /
• v.4 no.2
• /
• pp.87-99
• /
• 1999

We developed a Part Information Brokering System, namely PIBS, that can inter-link part suppliers and buyers using WWW(World Wide Web). The major contribution of this paper is that the system provides a 3D collaboration environment that enables multiple users to share the same workspace where they cooperatively manipulate part features in real time. PIBS carries out two major functions: one to maintain part data in a part library, and the other to provide the means for the clients to cooperate. An object-oriented database is used for the part library which stores apart information registered by part suppliers. VRML(Virtual Reality Modeling Language) the ISO standard for 3D visualization on WWW, is used to represent 3D part models. Several Java programs have been implemented to support synchronous and a synchronous communication of the 3D models. Once the users are interconnected through the system, a user's manipulation of part objects is transparently and instantaneously transmitted to the others. This means that all the participants can are the same view and movement of the 3D part models. Since the system is developed using a Java applet-server architecture, it requires no additional software other than standard web browser. The prototype system has been successfully implemented, and demonstrated its applicability to virtual 3D part information sharing.

• Proceedings of the Korea Concrete Institute Conference
• /
• 2009.05a
• /
• pp.43-44
• /
• 2009

3D bridge information models for a concrete box-girder bridge were built and a typical application example for the estimation linked with the 3D models was proposed. 3D geometry objects with PBS (product breakdown structure) were constructed using standard classification codes and 2D drawings. Information on the amount of materials according to the classification codes was utilized for the estimation system. A process for the estimation of concrete and rebars was developed and verified through the pilot test. Optimization of rebar fabrication was suggested as an additional function and was utilized for the cost reduction and standard fabrication.

• Journal of Broadcast Engineering
• /
• v.27 no.2
• /
• pp.207-215
• /
• 2022

The latest volumetric technology's high geometrical accuracy and realism ensure a high degree of correspondence between the real object and the captured 3D model. Nevertheless, since the 3D model obtained in this way constitutes a sequence as a completely independent 3D model between frames, the consistency of the model surface structure (geometry) is not guaranteed for every frame, and the density of vertices is very high. It can be seen that the interconnection node (Edge) becomes very complicated. 3D models created using this technology are inherently different from models created in movie or video game production pipelines and are not suitable for direct use in applications such as real-time rendering, animation and simulation, and compression. In contrast, our method achieves consistency in the quality of the volumetric 3D model sequence by linking re-meshing, which ensures high consistency of the 3D model surface structure between frames and the gradual deformation and texture transfer through correspondence and matching of non-rigid surfaces. And It maintains the consistency of volumetric 3D model sequence quality and provides post-processing automation.

• Journal of Korean Institute of Industrial Engineers
• /
• v.40 no.3
• /
• pp.342-357
• /
• 2014

The present study analyzes the relationship between research and development (R&D) inputs and performance of a national technology innovation R&D program using successive binary Logistic regression models based on a typical R&D logic model. In particular, this study focuses on to answer the following three main questions; (1) "To what extent, do the R&D inputs have an effect on the performance creation?"; (2) "Is an obvious relationship verified between the immediate predecessor and its successor performance?"; and (3) "Is there a difference in the performance creation between R&D government subsidy recipient types and between R&D collaboration types?" Methodologically, binary Logistic regression models are established successively considering the "Success-Failure" binary data characteristic regarding the performance creation. An empirical analysis is presented analyzing the sample n = 2,178 R&D projects completed. This study's major findings are as follows. First, the R&D inputs have a statistically significant relationship only with the short-term, technical output, "Patent Registration." Second, strong dependencies are identified between the immediate predecessor and its successor performance. Third, the success probability of the performance creation is statistically significantly different between the R&D types aforementioned. Specifically, compared with "Large Company", "Small and Medium-Sized Enterprise (SMS)" shows a greater success probability of "Sales" and "New Employment." Meanwhile, "R&D Collaboration" achieves a larger success probability of "Patent Registration" and "Sales."

• IEMEK Journal of Embedded Systems and Applications
• /
• v.18 no.5
• /
• pp.203-208
• /
• 2023

Recently, with the growing interest in the field of autonomous driving, many researchers have been focusing on developing autonomous driving software platforms. In particular, we have concentrated on developing 3D object detection models that can improve real-time performance. In this paper, we introduce a self-constructed 3D LiDAR dataset specific to domestic environments and propose a VariFocal-based CenterPoint for the 3D object detection model, with improved performance over the previous models. Furthermore, we present experimental results comparing the performance of the 3D object detection modules using our self-built and public dataset. As the results show, our model, which was trained on a large amount of self-constructed dataset, successfully solves the issue of failing to detect large vehicles and small objects such as motorcycles and pedestrians, which the previous models had difficulty detecting. Consequently, the proposed model shows a performance improvement of about 1.0 mAP over the previous model.

• Journal of Computational Design and Engineering
• /
• v.2 no.1
• /
• pp.47-54
• /
• 2015

In recent years, 3D printers have become popular as a means of outputting geometries designed on CAD or 3D graphics systems. However, the complex user interfaces of standard 3D software can make it difficult for ordinary consumers to design their own objects. Furthermore, models designed on 3D graphics software often have geometrical problems that make them impossible to output on a 3D printer. We propose a novel AR (augmented reality) 3D modeling system with an air-spray like interface. We also propose a new data structure (octet voxel) for representing designed models in such a way that the model is guaranteed to be a complete solid. The target shape is based on a regular polyhedron, and the octet voxel representation is suitable for designing geometrical objects having the same symmetries as the base regular polyhedron. Finally, we conducted a user test and confirmed that users can intuitively design their own ornaments in a short time with a simple user interface.

• Proceedings of the KIEE Conference
• /
• 2005.07d
• /
• pp.2897-2899
• /
• 2005

Finite element method (FEM) has been widely used as a useful numerical method that can analyze complex engineering problems in electro-magnetics, mechanics, and others. CEMTool, which is similar to MATLAR, is a command style design and analyzing package for scientific and technological algorithm and a matrix based computation language. In this paper, we present new 3D FEM package in CEMTool environment. In contrast to the existing CEMTool 2D FEM package and MATLAB PDE (Partial Differential Equation) Toolbox, our proposed 3D FEM package can deal with complex 3D models, not a cross-section of 3D models. Consequently, with our new 3D FEM toolbox, we can analyze more diverse engineering Problems which the existing CEMTool 2D FEM package or MATLAB PDE Toolbox can not solve.