Filters








32,699 Hits in 4.2 sec

GPU-based shape from silhouettes

Sofiane Yous, Hamid Laga, Masatsugu Kidode, Kunihiro Chihara
2007 Proceedings of the 5th international conference on Computer graphics and interactive techniques in Australia and Southeast Asia - GRAPHITE '07  
In this paper, we present a new method for surface-based shape reconstruction from a set of silhouette images.  ...  We propose to project the viewing cones from all viewpoints to the 3D space and compute the intersections that represent the vertices of the Visual Hull (VH).  ...  At each rendering Contour images Extracted edges Closeup views Figure 5 : Viewing edge extraction steps. step, appropriate textures are attached as input(s) and output(s).  ... 
doi:10.1145/1321261.1321274 dblp:conf/graphite/YousLKC07 fatcat:jvcvmv7dkza7rlcwaqbniyevj4

A Minimalist Approach to Facial Reconstruction [article]

M. Gotla, Z. Huang
1999 Eurographics State of the Art Reports  
By applying cognitive and anatomical heuristics, we show that a realistic face model can be reconstructed from a generic face model using only one or two real facial images and minimal user interaction  ...  But if only the front view image is available for a face, e.g., from a web page, ID or passport photo, the 3D-reconstruction result is still very distinguishable.  ...  In Figure 3 , the reconstruction results of the top row are obtained using only one image -the front view of the face. The results in bottom row are obtained using two images -front and profile.  ... 
doi:10.2312/egs.19991027 fatcat:avqtdfkynfe4zfs6rb56uh3a5u

A Fast Surface-Based Visual Hull Reconstruction

Sofiane Yous, Masatsugu Kidode
2007 IAPR International Workshop on Machine Vision Applications  
Figure 6 shows the extracted edges using four silhouette images of a bunny taken from 4 viewpoints.  ...  Usually, the viewing edges are extracted by projecting each ray from each view to the silhouettes of all other views and find the line segments that intersect all silhouettes.  ... 
dblp:conf/mva/YousK07 fatcat:bfnqm3yy3na73eu55czgqqor6i

Line Drawings from 3D Models: A Tutorial

Pierre Bénard, Aaron Hertzmann
2019 Foundations and Trends in Computer Graphics and Vision  
This tutorial describes the geometry and algorithms for generating line drawings from 3D models, focusing on occluding contours.  ...  The geometry of occluding contours on meshes and on smooth surfaces is described in detail, together with algorithms for extracting contours, computing their visibility, and creating stylized renderings  ...  In addition to speed, image-space lines have the advantage that they automatically remove clutter as a function of image-space line density, although, like all image-based methods, they potentially lose  ... 
doi:10.1561/0600000075 fatcat:vfuxfjozhbgjrkcburw4geuvrq

Line Drawings from 3D Models [article]

Pierre Bénard, Aaron Hertzmann
2019 arXiv   pre-print
This tutorial describes the geometry and algorithms for generating line drawings from 3D models, focusing on occluding contours.  ...  The geometry of occluding contours on meshes and on smooth surfaces is described in detail, together with algorithms for extracting contours, computing their visibility, and creating stylized renderings  ...  First, when a closed mesh with outward-facing normals is viewed from the outside, the back-faces must all be occluded by front-faces.  ... 
arXiv:1810.01175v2 fatcat:23jyjtvbfjem7eq46mo2rbct4e

DeepProfile: Accurate Under-the-Clothes Body Profile Estimation

Shufang Lu, Funan Lu, Xufeng Shou, Shuaiyin Zhu
2022 Applied Sciences  
In this paper, we created and labeled an under-the-clothes human body contour keypoint dataset; we utilized a convolutional neural network (CNN) to extract the contour keypoints, then combined them with  ...  Image-based human body profile estimation can be regarded as a fine-grained semantic segmentation problem, which is typically used to locate objects and boundaries in images.  ...  There was a small number of front-view images in which the subject was partly obscured by something. In the side-view images, the faces of all the subjects were facing right.  ... 
doi:10.3390/app12042220 fatcat:6iroqgdr2nemvjudwd3kz52hie

Snaxels on a plane

Kevin Karsch, John C. Hart
2011 Proceedings of the ACM SIGGRAPH/Eurographics Symposium on Non-Photorealistic Animation and Rendering - NPAR '11  
The snaxel active contours can also track contours as the mesh animates, and frame-to-frame correspondences between snaxels lead to a new method to convert the moving contours on a 3-D animated mesh into  ...  While many algorithms exist for tracing various contours for illustrating a meshed object, few algorithms organize these contours into region-bounding closed loops.  ...  As shown previously ], using normals interpolated from vertex normals to extract a smooth visual contour generator over mesh faces instead of edges [Hertzmann and Zorin 2000] , can lead to a silhouette  ... 
doi:10.1145/2024676.2024683 dblp:conf/npar/KarschH11 fatcat:mps2vgdmanb3paqhccqcil6awi

Photorealistic 3D face modeling on a smartphone

Won Beom Lee, Man Hee Lee, In Kyu Park
2011 CVPR 2011 WORKSHOPS  
Major features of human face such as eyes, nose, lip, cheek, chin, and profile boundary are extracted automatically from the front and profile images, in which ACM (active contour model) and deformable  ...  Skin texture map is created from the input image, which is mapped on the deformed 3D face model.  ...  Akimoto, Park, and Ansari et al. automatically extracted the facial features from the frontal and profile images and then generated a customized 3D face model by deforming the generic model to fit to these  ... 
doi:10.1109/cvprw.2011.5981841 dblp:conf/cvpr/LeeLP11 fatcat:me52drfudnbdrovz3tebxgyis4

Three-Dimensional Face Reconstruction Using Multi-View-Based Bilinear Model

Liang Tian, Jing Liu, Wei Guo
2019 Sensors  
Here, we propose a novel facial reconstruction framework that accurately extracts the 3D shapes and poses of faces from images captured at multi-views.  ...  multi-view images.  ...  It is shown how the feature prior and texture constraints from the multi-view images facilitate the extraction of a 3D face mesh.  ... 
doi:10.3390/s19030459 fatcat:vlrhchyjvvd43i4e4i4gprjse4

Face Profile View Retrieval Using Time of Flight Camera Image Analysis [chapter]

Piotr Bratoszewski, Andrzej Czyżewski
2015 Lecture Notes in Computer Science  
Method for profile view retrieving of the human face is presented. The depth data from the 3D camera is taken as an input.  ...  Knowing the locations of keypoints and having the depth data the contour of the user's face a profile retrieval is attempted.  ...  Their approach uses high resolution face profile images constructed from low resolution videos from a side view camera.  ... 
doi:10.1007/978-3-319-19941-2_16 fatcat:7hww3rwfe5cinabywerwstdoqi

Snake head boundary extraction using global and local energy minimisation

S.R. Gunn, M.S. Nixon
1996 Proceedings of 13th International Conference on Pattern Recognition  
The new technique employing dynamic programming is deployed to extract the inner face boundary, along with a conventional normaldriven technique to extract the outer face boundary.  ...  Application to a database of 75 subjects showed that the outer contour was extracted successfully for 96% of the subjects and the inner contour was successful for 82%.  ...  As with the previous approaches, our new approach uses front view face images.  ... 
doi:10.1109/icpr.1996.546890 dblp:conf/icpr/GunnN96 fatcat:mzk7jbmajzdqdncini4mrov6ca

Automatic Facial Spots and Acnes Detection System

Chuan-Yu Chang, Heng-Yi Liao
2013 Journal of Cosmetics Dermatological Sciences and Applications  
The system first locates the facial region from the input image. Then, the shapes of faces were recognized using a contour descriptor.  ...  The facial features are extracted to define regions of interest and an image segment method is used to extract potential defect.  ...  There are 54 face images taken from 3 females and 15 males in 3 views. Face View Classification Result To demonstrate the capability of the face view classification, the two databases are used.  ... 
doi:10.4236/jcdsa.2013.31a006 fatcat:ihftwii7zzbn5g7kctj2uvar3m

Isoradius Contours: New Representations and Techniques for 3D Face Registration and Matching

Nick Pears, Tom Heseltine
2006 Third International Symposium on 3D Data Processing, Visualization, and Transmission (3DPVT'06)  
We propose a technique for 3D face registration and matching using a novel representation called "isoradius contours".  ...  A 3D face representation contains many isoradius contours with different radii and the first major benefit of the technique is that the shape of the contours is independent of the facial pose, due to the  ...  Figure 2 :Figure 3 : 23 Irad contours on 3D image 00179-08: front view Irad contours on 3D image 00179-08: profile view Figure 4 :Figure 5 :Figure 6 : 2 Figure 7 :Figure 8 : 456278 Marked subject frontal  ... 
doi:10.1109/3dpvt.2006.85 dblp:conf/3dpvt/PearsH06 fatcat:otecogahl5hvvimi6z6yhv7eyq

Shadow Volume Generation from Free Form Surfaces [chapter]

Gregory Heflin, Gershon Elber
1993 Communicating with Virtual Worlds  
In this paper, an approach is presented that allows one to compute shadow v olumes directly from free form models and exploit them for the generation of shadows using a Z-bu er based renderer.  ...  Then g(t) = ( n 1 (u(t) v (t)) Ṽ )(ñ 2 (r(t) s (t)) Ṽ ) (5) is positive if both normals are front facing or back facing, and is negative if one is front facing and the other is back facing.  ...  Figure 4 : 4 View is from the light source. Contours are shown in bold.  ... 
doi:10.1007/978-4-431-68456-5_10 fatcat:kz652ynqpnbllmrp3sh3xnoxgy

Model-Based Generation of Realistic 3D Full Body Avatars from Uncalibrated Multi-view Photographs [chapter]

Nicholas Michael, Andreas Lanitis
2014 IFIP Advances in Information and Communication Technology  
The proposed method utilizes a statistical model of human 3D shape and a multi-view statistical 2D shape model of its corresponding silhouettes.  ...  Therefore, we present here a novel modelbased technique that is capable of real-time generation of personalized full-body 3D avatars from orthogonal photographs.  ...  Silhouette Multi-view ASM We project the 3D range scans of the training set to four orthogonal views (front, left, right and back) extracting the corresponding 2D silhouette contour in each view.  ... 
doi:10.1007/978-3-662-44654-6_35 fatcat:5tdvjwmxr5eavleelc2lr4odii
« Previous Showing results 1 — 15 out of 32,699 results