• Journal of the HCI Society of Korea
• /
• v.1 no.1
• /
• pp.53-61
• /
• 2006

In this paper, we propose a novel method for 3D reconstruction of an indoor scene using a multi-view camera. Until now, numerous disparity estimation algorithms have been developed with their own pros and cons. Thus, we may be given various sorts of depth images. In this paper, we deal with the generation of a 3D surface using several 3D point clouds acquired from a generic multi-view camera. Firstly, a 3D point cloud is estimated based on spatio-temporal property of several 3D point clouds. Secondly, the evaluated 3D point clouds, acquired from two viewpoints, are projected onto the same image plane to find correspondences, and registration is conducted through minimizing errors. Finally, a surface is created by fine-tuning 3D coordinates of point clouds, acquired from several viewpoints. The proposed method reduces the computational complexity by searching for corresponding points in 2D image plane, and is carried out effectively even if the precision of 3D point cloud is relatively low by exploiting the correlation with the neighborhood. Furthermore, it is possible to reconstruct an indoor environment by depth and color images on several position by using the multi-view camera. The reconstructed model can be adopted for interaction with as well as navigation in a virtual environment, and Mediated Reality (MR) applications.

• Current Optics and Photonics
• /
• v.5 no.6
• /
• pp.680-685
• /
• 2021

We propose a technique that reconstructs a hologram whose pixel number is greater than the pixel numbers of a conventional image sensor. The pixel numbers of the hologram recorded by optical scanning holography (OSH) increases as the scan area becomes larger. The reconstruction time also increases drastically as the size of the hologram increases. The holographic information of a three-dimensional (3D) scene is distributed throughout the recorded hologram; this makes the simple divide-and-stitch approach fail. We propose a technique that reconstructs the hologram without loss of holographic information. First, we record the hologram of a 3D scene using OSH. Second, we segment the hologram into sub-holograms that contain complete holographic information. Third, we reconstruct the sub-holograms simultaneously. Finally, we rearrange the reconstructions of the sub-holograms.

• Proceedings of the IEEK Conference
• /
• 2002.07b
• /
• pp.1145-1148
• /
• 2002

This paper presents a new method which is calculated to use only three vanishing points in order to compute the dimensions of object and its pose from a single image of perspective projection taken by a camera and the problem of recovering 3D models from three vanishing points of box scene. Our approach is to compute only three vanishing points without this information such as the focal length, rotation matrix, and translation from images in the case of perspective projection. We assume that the object can be modeled as a linear function of a dimension vector ν. The input of reconstruction is a set of correspondences between features in the model and features in the image. To minimize each the dimensions of the parameterized models, this reconstruction of optimization can be solved by the standard nonlinear optimization techniques with a multi-start method which generates multiple starting points for the optimizer by sampling the parameter space uniformly.

• Journal of the Optical Society of Korea
• /
• v.12 no.3
• /
• pp.131-135
• /
• 2008

Computational integral imaging(CII) has the advantage of generating the volumetric information of the 3D scene without optical devices. However, the reconstruction process of CII requires increasingly larger sizes of reconstructed images and then the computational cost increases as the distance between the lenslet array and the reconstructed output plane increases. In this paper, to overcome this problem, we propose a novel CII method using a depth conversion technique. The proposed method can move a far 3D object near the lenslet array and reduce the computational cost dramatically. To show the usefulness of the proposed method, we carry out the preliminary experiment and its results are presented.

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.153.5-153
• /
• 2001

This paper presents a simple modeling system that constructs 3D models from an indoor cylindrical environment map using all of the available geometry of the interior structure such as vertical and horizontal lines and parallel and perpendicular planes. The indoor scene abstract model is created through this system and the navigation through the process of 3D reconstruction. This system first automatically detects the vanishing points in a cylindrical environment map from parallel lines and planes, and determines the indoor scene topology previously defined using this information. The determined topology enables he user intervention UI simply construct a 3D model by using the photogrammetry. The modeling system can be ...

• Journal of KIISE:Software and Applications
• /
• v.31 no.11
• /
• pp.1543-1550
• /
• 2004

Modeling from images is a cost-effective means of obtaining 3D geometric models. These models can be effectively constructed from classical Structure from Motion algorithm. However, it's too difficult to reconstruct whole scenes using SFM method since general sites contain a very complex shapes and brilliant colours. To overcome this difficulty, the current paper proposes a new reconstruction method based on a moving Planar mirror. We devise the mirror posture instead of scene itself as a cue for reconstructing the geometry That implies that the geometric cues are inserted into the scene by compulsion. With this method, we can obtain the geometric details regardless of the scene complexity. For this purpose, we first capture image sequences through the moving mirror containing the interested scene, and then calibrate the camera through the mirror's posture. Since the calibration results are still inaccurate due to the detection error, the camera pose is revised using frame-correspondence of the comer points that are easily obtained using the initial camera posture. Finally, 3D information is computed from a set of calibrated image sequences. We validate our approach with a set of experiments on some complex objects.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.11a
• /
• pp.646-647
• /
• 2022

Multiview stereo (MVS) 3D reconstruction of a scene from images is a fundamental computer vision problem that has been thoroughly researched in recent times. Traditionally, MVS approaches create dense correspondences by constructing regularizations and hand-crafted similarity metrics. Although these techniques have achieved excellent results in the best Lambertian conditions, traditional MVS algorithms still contain a lot of artifacts. Therefore, in this study, we suggest using a transformer network to accelerate the MVS reconstruction. The network is based on a transformer model and can extract dense features with 3D consistency and global context, which are necessary to provide accurate matching for MVS.

• Journal of Internet Computing and Services
• /
• v.10 no.4
• /
• pp.25-33
• /
• 2009

In the paper, we proposed a new system to retrieve 3D wrinkle data based on stereo images. Usually, 3D reconstruction based on stereo images or video is very popular and it is the research focus, which has been applied for culture heritage, building and other scene. The target is object measurement, the scene depth calculation and 3D data obtained. There are several challenges in our research. First, it is hard to take the full information wrinkle images by cameras because of light influence, skin with non-rigid object and camera performance. We design a particular computer vision system to take winkle images with a long length camera lens. Second, it is difficult to get the dense stereo data because of the hard skin texture image segmentation and corner detection. We focus on semi-dense stereo matching algorithm for the wrinkle depth. Compared with the 3D scanner, our system is much cheaper and compared with the physical modeling based method, our system is more flexible with high performance.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.42 no.2 s.302
• /
• pp.113-120
• /
• 2005

3D reconstruction of a scene structure from un-calibrated image sequences has been long one of the central problems in computer vision. For 3D reconstruction in Euclidean space, projective reconstruction, which is classified into the merging method and the factorization, is needed as a preceding step. By calculating all camera projection matrices and structures at the same time, the factorization method suffers less from dia and error accumulation than the merging. However, the factorization is hard to analyze precisely long sequences because it is based on the assumption that all correspondences must remain in all views from the first frame to the last. This paper presents a new projective reconstruction method for recovery of 3D structure over long sequences. We break a full sequence into sub-sequences based on a quantitative measure considering the number of matching points between frames, the homography error, and the distribution of matching points on the frame. All of the projective reconstructions of sub-sequences are registered into the same coordinate frame for a complete description of the scene. no experimental results showed that the proposed method can recover more precise 3D structure than the merging method.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.30 no.8C
• /
• pp.775-782
• /
• 2005

In case of 3D reconstruction from dense data in uncalibrated sequence images, we encounter with the problem for searching many correspondences and the computational costs. In this paper, we propose a key frame selection method from uncalibrated images and the effective 3D reconstruction method using the key frames. Namely, it can be performed on smaller number of views in the image sequence. We extract correspondences from selected key frames in image sequences. From the extracted correspondences, camera calibration process will be done. We use the edge image to fed dense correspondences between selected key frames. The method we propose to find dense correspondences can be used for recovering the 3D structure of the scene more efficiently.