• International Journal of Internet, Broadcasting and Communication
• /
• v.16 no.2
• /
• pp.81-86
• /
• 2024

Advancements in generative AI are reshaping graphic and 3D content design landscapes, where AI not only enriches graphic design but extends its reach to 3D content creation. Though 3D texture mapping through AI is advancing, AI-generated 3D modeling technology in this realm remains nascent. This paper presents AI 2D image-driven 3D modeling techniques, assessing their viability in 3D content design by scrutinizing various algorithms. Initially, four OBJ model-exporting AI algorithms are screened, and two are further evaluated. Results indicate that while AI-generated 3D models may not be directly usable, they effectively capture reference object structures, offering substantial time savings and enhanced design efficiency through manual refinements. This endeavor pioneers new avenues for 3D content creators, anticipating a dynamic fusion of AI and 3D design.

• The Journal of Korea Robotics Society
• /
• v.10 no.1
• /
• pp.33-41
• /
• 2015

Recent developments in robotics and intelligent vehicle area, bring interests of people in an autonomous driving ability and advanced driving assistance system. Especially fully automatic parking ability is one of the key issues of intelligent vehicles, and accurate parked vehicles detection is essential for this issue. In previous researches, many types of sensors are used for detecting vehicles, 2D LiDAR is popular since it offers accurate range information without preprocessing. The L shape feature is most popular 2D feature for vehicle detection, however it has an ambiguity on different objects such as building, bushes and this occurs misdetection problem. Therefore we propose the accurate vehicle detection method by using a 3D complete vehicle model in 3D point clouds acquired from front inclined 2D LiDAR. The proposed method is decomposed into two steps: vehicle candidate extraction, vehicle detection. By combination of L shape feature and point clouds segmentation, we extract the objects which are highly related to vehicles and apply 3D model to detect vehicles accurately. The method guarantees high detection performance and gives plentiful information for autonomous parking. To evaluate the method, we use various parking situation in complex urban scene data. Experimental results shows the qualitative and quantitative performance efficiently.

• The Transactions of the Korea Information Processing Society
• /
• v.13 no.7
• /
• pp.335-347
• /
• 2024

Open-vocabulary 3D point cloud instance segmentation (OV-3DIS) is a challenging visual task to segment a 3D scene point cloud into object instances of both base and novel classes. In this paper, we propose a novel model Open3DME for OV-3DIS to address important design issues and overcome limitations of the existing approaches. First, in order to improve the quality of class-agnostic 3D masks, our model makes use of T3DIS, an advanced Transformer-based 3D point cloud instance segmentation model, as mask proposal module. Second, in order to obtain semantically text-aligned visual features of each point cloud segment, our model extracts both 2D and 3D features from the point cloud and the corresponding multi-view RGB images by using pretrained CLIP and OpenSeg encoders respectively. Last, to effectively make use of both 2D and 3D visual features of each point cloud segment during label assignment, our model adopts a unique feature ensemble method. To validate our model, we conducted both quantitative and qualitative experiments on ScanNet-V2 benchmark dataset, demonstrating significant performance gains.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 1996.06b
• /
• pp.18-21
• /
• 1996

In the field of Virtual Reality, many studies have been reported. Especially, there are many studies on generating virtual creatures on computer systems. In this paper we propose an algorithm to automatically generate 3D fish models from 2D images which are printed in illustrated books, pictures or handwritings. At first, 2D fish images are captured by means of image scanner. Next, the fish image is separated from background and segmented to several parts such as body, anal fin, dorsal fin, ectoral fin and ventral fin using the proposed method“Active Balloon model”. After that, users choose front view model and top view model among six samples, respectively. 3D model is automatically generated from separated body, fins and the above two view models. The number of patches is decreased without any influence on the accuracy of the generated 3D model to reduce the time cost when texture mapping is applied. Finally, we can get any kinds of 3D fish models.

• Korean Journal of Computational Design and Engineering
• /
• v.6 no.2
• /
• pp.133-139
• /
• 2001

Since a great number of 2D engineering drawings are being used in industry and at the same time 3D CAD becomes popular in recent years, we need to reconstruct 3D CAD models from 2D legacy drawings. In this thesis, a combination of a feature recognition method and an expert system is suggested for the 3D solid model reconstruction. Modeling primitives of 3D CAD systems are recognized and constructed by using the pattern matching technique of the features modeling. Additional information for the 3D model reconstruction can be generated by extracting symbols or text entities which are related to form entities. For complex and indefinite cases which cannot be solved by the process of feature recognition, an expert system with a rule base has been used for decision-making. A 3D reconstruction system which recognizes 2D DXF drawing files has been implemented where models composed with protrusions, holes, and cutouts can be handled.

• Journal of Biomedical Engineering Research
• /
• v.15 no.3
• /
• pp.325-332
• /
• 1994

In this paper, a method is proposed to detect the endocardial contour of the left ventricle using the Hough transform with a weighted model and edge information from the 2D echocardiogram. The implementation of this method is as follows: first, an approximate model detection algorithm was implemented in order to detect the approximate endocardium model and the model center, then we constructed a weighted model with the detected model. Next, we found automatically the cavity center of the left ventricle performing the Hough transform which used the weighted model, and then we detected the endocardial contour using weighted model and edge image.

• Proceedings of the IEEK Conference
• /
• 2006.06a
• /
• pp.335-336
• /
• 2006

In this paper, we propose an active shape image segmentation method for three-dimensional(3-D) medical images using a generation method of the 3-D shape model. The proposed method generates the shape model using a distance transform and a tetrahedron method for landmarking. After generating the 3-D model, we extend the training and segmentation processes of 2-D active shape model(ASM) and improve the searching process. The proposed method provides comparative results to 2-D ASM, region-based or contour-based methods. Experimental results demonstrate that this algorithm is effective for a semi-automatic segmentation method of 3-D medical images.

• Journal of the Korean Society for Railway
• /
• v.14 no.1
• /
• pp.11-18
• /
• 2011

This paper proposed a new 2-D multi-body dynamic modeling technique to analyze overriding behaviors taking place during train collision. This dynamic model is composed of nonlinear springs, dampers and masses by considering the deformable characteristics of carbodies as well as energy absorbing structures and components. By solving this dynamic model for rollingstock, energy absorbing capacities of collision elements, accelerations of passenger sections, impact forces applied to interconnecting devices, and overriding displacements can be well estimated. For a case study, we chose KHST (Korean High Speed Train), obtained crush characteristic data of each carbody section from 3-D finite element analysis, and established a 2-D multi-body dynamic model. This 2-D dynamic model was simulated under the train-to-train collision scenarios, and evaluated with 3-D virtual testing model. It was founded from the simulation results that this 2-D dynamic model could well predict overriding behaviors, and the modeling technique of carbody deformation was very important in overriding estimation.

• Journal of Technology Innovation
• /
• v.31 no.2
• /
• pp.83-110
• /
• 2023

This study deals with the process by which 'Social problem-solving R&D model' is established in Korean society through the evolution of the government's R&D program. We will examine the process by which a 'Social problem-solving R&D model', which was not present in companies, science and technology communities, and civil society, is formed through government R&D programs. To this end, we present a conceptual framework to analyze the process of co-evolution of 'Social problem solving R&D model', the organizational community that supports and implements it, and the institutions that supports the new model. In the synthesis, policy measures to enhance 'Social problem-solving R&D model' are dealt with.

• Journal of Korea Water Resources Association
• /
• v.54 no.5
• /
• pp.301-309
• /
• 2021

In this study, we tried to improve the performance of the existing U-net-based deep learning rainfall prediction model, which can weaken the meaning of time series order. For this, ConvLSTM2D U-Net structure model considering temporal consistency of data was applied, and we evaluated accuracy of the ConvLSTM2D U-Net model using a RainNet model and an extrapolation-based advection model. In addition, we tried to improve the uncertainty in the model training process by performing learning not only with a single model but also with 10 ensemble models. The trained neural network rainfall prediction model was optimized to generate 10-minute advance prediction data using four consecutive data of the past 30 minutes from the present. The results of deep learning rainfall prediction models are difficult to identify schematically distinct differences, but with ConvLSTM2D U-Net, the magnitude of the prediction error is the smallest and the location of rainfall is relatively accurate. In particular, the ensemble ConvLSTM2D U-Net showed high CSI, low MAE, and a narrow error range, and predicted rainfall more accurately and stable prediction performance than other models. However, the prediction performance for a specific point was very low compared to the prediction performance for the entire area, and the deep learning rainfall prediction model also had limitations. Through this study, it was confirmed that the ConvLSTM2D U-Net neural network structure to account for the change of time could increase the prediction accuracy, but there is still a limitation of the convolution deep neural network model due to spatial smoothing in the strong rainfall region or detailed rainfall prediction.