• International journal of advanced smart convergence
• /
• v.12 no.4
• /
• pp.142-146
• /
• 2023

We present a method for generating 3D structures and rendering objects by combining VAE (Variational Autoencoder) and GAN (Generative Adversarial Network). This approach focuses on generating and rendering 3D models with improved quality using residual learning as the learning method for the encoder. We deep stack the encoder layers to accurately reflect the features of the image and apply residual blocks to solve the problems of deep layers to improve the encoder performance. This solves the problems of gradient vanishing and exploding, which are problems when constructing a deep neural network, and creates a 3D model of improved quality. To accurately extract image features, we construct deep layers of the encoder model and apply the residual function to learning to model with more detailed information. The generated model has more detailed voxels for more accurate representation, is rendered by adding materials and lighting, and is finally converted into a mesh model. 3D models have excellent visual quality and accuracy, making them useful in various fields such as virtual reality, game development, and metaverse.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.25 no.6_1
• /
• pp.545-555
• /
• 2007

3D buildings are the most prominent feature comprising urban scene. A few of mega-cities in the globe are virtually reconstructed in photo-realistic 3D models, which becomes accessible by the public through the state-of-the-art online mapping services. A lot of research efforts have been made to develop automatic reconstruction technique of large-scale 3D building models from remotely sensed data. However, existing methods still produce irregular building polygons due to errors induced partly by uncalibrated sensor system, scene complexity and partly inappropriate sensor resolution to observed object scales. Thus, a geometric regularization technique is urgently required to rectify such irregular building polygons that are quickly captured from low sensory data. This paper aims to develop a new method for regularizing noise building outlines extracted from airborne LiDAR data, and to evaluate its performance in comparison with existing methods. These include Douglas-Peucker's polyline simplication, total least-squared adjustment, model hypothesis-verification, and rule-based rectification. Based on Minimum Description Length (MDL) principal, a new objective function, Geometric Minimum Description Length (GMDL), to regularize geometric noises is introduced to enhance the repetition of identical line directionality, regular angle transition and to minimize the number of vertices used. After generating hypothetical regularized models, a global optimum of the geometric regularity is achieved by verifying the entire solution space. A comparative evaluation of the proposed geometric regulator is conducted using both simulated and real building vectors with various levels of noise. The results show that the GMDL outperforms the selected existing algorithms at the most of noise levels.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.21 no.3
• /
• pp.229-236
• /
• 2003

This study recorded an object using a digital video recorder, and then tried to estimate 3-D positional information and to reconstruct an image. Firstly, the accuracy of measurement results from a video recorder was evaluated and tested for an applicability, then it applied to a real object to construct 3-D digital model. This study assumed that there is no lens distortion in a video recorder, and all bundles should precisely pass through the projection center of a lens. The image size for orientations is determined by the size of CCD chip and the number of pixels. The average squared error from the result by a digital video recorder and that by triangular survey from 1-second theodolite shows 0.0173m error in x,y coordinates. Without knowing the accurate information on the lens distortion and the coordinates of the projection center, this study reasonably produces acceptable results in the reconstruction of 3-D model. In consequence, this study found that the image from a digital video camera can be reconstructed 3-D model only from the information on a camera type.

• New Physics: Sae Mulli
• /
• v.68 no.11
• /
• pp.1275-1280
• /
• 2018

We demonstrated transmission direct phase-measuring deflectometry (DPMD) with a specular phase object having discontinuous surfaces by using two displays and a two-dimensional array detector for display and by recording the distorted fringe patterns. Three-dimensional (3D) information was obtained by calculating the height map directly from the phase information. We developed a mathematical model of the phase-height relationship in transmission DPMD. Unlike normal transmission deflectometry, this method supports height measurement directly from the phase. Compared with other 3D measurement techniques such as interferometry, this method has the advantages of being inexpensive and easy to implement.

• The KIPS Transactions:PartB
• /
• v.12B no.3 s.99
• /
• pp.247-256
• /
• 2005

One of the classic research problems in computer vision is that of stereo, i.e., the reconstruction of three dimensional shape from two or more images. This paper deals with the problem of extracting depth information of non-rigid dynamic 3D scenes from general 2D video sequences taken by monocular camera, such as movies, documentaries, and dramas. Depth of the blocks are extracted from the resultant block motions throughout following two steps: (i) calculation of global parameters concerned with camera translations and focal length using the locations of blocks and their motions, (ii) calculation of each block depth relative to average image depth using the global parameters and the location of the block and its motion, Both singular and non-singular cases are experimented with various video sequences. The resultant relative depths and ego-motion object shapes are virtually identical to human vision.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.33 no.6
• /
• pp.2509-2518
• /
• 2013

Augmented reality (AR) technology visualizes a real type object that cannot simulate in virtual reality technology by overlapping a virtual object and real object in a computer system. This study suggests a methodology and prototype system for applying AR system to rebar distribution work in a civil engineering project. Rebar work in civil engineering project is a representative activity that is progressed by empirical approach of skilled labor rather than formalized manual. AR technology improves the constructability of rebar work because AR tool can identify missing rebars and different rebars comparing with the drawings. AR system developed in this study can enhance the understanding of rebar work using 3D modeling with real image of construction site and save construction cost by reducing reconstruction work.

• Journal of Broadcast Engineering
• /
• v.19 no.5
• /
• pp.677-686
• /
• 2014

This paper analyzes the 3D accuracy of stereo images captured from a mobile phone. For 3D accuracy evaluation, we have compared the accuracy result according to the amount of the convergence angle. In order to calculate the 3D model space coordinate of control points, we perform inner orientation, distortion correction and image geometry estimation. And the quantitative 3D accuracy was evaluated by transforming the 3D model space coordinate into the 3D object space coordinate. The result showed that relatively precise 3D information is generated in more than $17^{\circ}$ convergence angle. Consequently, it is necessary to set up stereo model structure consisting adequate convergence angle as an measurement distance and a baseline distance for accurate 3D information generation. It is expected that the result would be used to stereoscopic 3D contents and 3D reconstruction from images captured by a mobile phone camera.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.26 no.7B
• /
• pp.957-967
• /
• 2001

본 논문에서는 모델 기반 지각적 그룹핑을 이용한 3차원 다면체의 인식 및 형상 복원에 관한 새로운 기법을 제안한다. 2차원 입력 영상과 여기에서 추출된 특징들의 3차원 특징을 거리 측정기를 이용하여 추출하여 인식 및 복원의 기본 특징으로 이용한다. 이 때, 모델의 3차원 기하학적 정보는 결정 트리 분류기에 의하여 학습되며 지각적 그룹핑은 이와 같은 모델 기반으로 이루어진다. 또한, 1차 그룹핑의 결과로 얻어진 3차원 직선 특징간의 관계는 Gestalt 그래프로 표현되며 이것의 부그래프 분할을 통하여 인식을 위한 후보 그룹이 생성된다. 마지막으로 각각의 후보 그룹은 3차원 모델과 정렬되어 가장 잘 부합되는 그룹을 인식 결과로 생성하게 된다. 그리고 정렬의 결과로서 2차원 텍스춰를 추출하여 3차원 모델에 매핑함으로써 실제적인 3차원 형상을 복원할 수 있다. 제안하는 알고리듬의 성능을 평가하기 위하여 불록 영상과 지형 모델 보드 영상에 대하여 실험을 수행하였다. 실험 결과, 모델 기반의 그룹핑 기법은 결과 그룹의 수를 상당히 감소시켰으며 또한 잡음과 가리워짐에 강건한 인식과 복원 결과가 얻어졌다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2016.06a
• /
• pp.309-312
• /
• 2016

최근 실세계에 존재하는 물체의 3차원 형상과 색상을 디지털화하는 3차원 객체 복원에 대한 관심이 날로 증가하고 있다. 3차원 객체 복원은 영상 획득, 영상 보정, 점군 획득, 반복적 점군 정합, 무리 조정, 3차원 모델 표현과 같은 단계를 거처 통합된 3차원 모델을 생성한다. 그 중 반복적 점군 정합 방법은 카메라 궤적의 초기 값을 획득하는 방법으로서 무리 조정 단계에서 전역 최적 값으로의 수렴을 보장하기 위해 중요한 단계이다. 기존의 반복적 점군 정합 (iterative closest points) 방법에서는 시간이 지남에 따라 누적된 궤적 오차 때문에 발생하는 객체 표류 문제가 발생한다. 본 논문에서는 이 문제를 해결하기 위해 색상 영상에서 SIFT 특징점을 획득하고 3차원 점군을 얻은 뒤 가중치를 부여함으로써 점 군 간의 더 정확한 정합을 수행한다. 실험결과에서 기존의 방법과 비교하여 제안하는 방법이 절대 궤적 오차 (absolute trajectory error)가 감소하는 것을 확인 했고 복원된 3차원 모델에서 객체 표류 현상이 줄어드는 것을 확인했다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2014.06a
• /
• pp.135-138
• /
• 2014

본 논문에서는 2대의 Kinect 카메라를 이용하여 실세계의 3차원 객체에 대한 복원을 수행하는 방법을 제안한다. 먼저 깊이 가중치가 추가된 계층적 결합형 양방향 필터를 이용하여 Kinect로부터 얻은 원본 깊이 영상을 보정한다. 그리고 카메라 캘리브레이션을 이용하여 카메라의 내부 파라미터와 외부 파라미터를 획득한다. 이를 이용해 3차원 워핑을 수행하여 각 시점의 데이터를 3차원 공간에 점군 모델로 복원하고 표면 모델링 방법을 이용하여 3차원 객체의 매끄러운 표면 모델을 생성한다. 실시간에 가까운 속도를 내기 위해서 계층적 결합형 양방향 필터와 3차원 워핑을 병렬 처리 프레임워크인 CUDA로 구현하여 고속화하였다. 실험을 통해 분리된 각 시점에서의 깊이 정보를 하나의 통합된 3차원 공간에 복원할 수 있었고 초당 5 fps의 속도로 동작하는 것을 확인하였다.