Modeling hair from multiple views

Wei, Yichen; Ofek, Eyal; Quan, Long; Shum, Heung-Yeung

doi:10.1145/1073204.1073267

Cited by 98 publications

(82 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In 2005,Wei et al [77] proposed an image-based approach to capture the hair geometry.In 2006,Bertails et al [78] proposed to use Kirchhoff's equations for elastic rods to describe the motion of hair strands.The equations are solved representing a hair strand as a piecewise helical rod.In the same year,Moon and Marshner [79] proposed an efficient physically based technique to simulate the multiple scattering of hair based on volumetric photon mapping. In 2008,Moon et al [80] improved their previous technique [81] and developed a much faster volumetric rendering method to compute multiple scattering solutions.…”

Section: Hairmentioning

confidence: 99%

Techniques of Facial Synthesis: A Comprehensive Literature Review

Chandra¹,

Karmakar²,

Hegadi³

2013

IJCA

View full text Add to dashboard Cite

As an important area of computer graphics, realistic image synthesis provides principles and techniques for creating realistic imagery through computations based upon models of real-world objects and behaviors. Facial expression is one of the primary communication means of the human. The ultimate goal of realistic image synthesis is to achieve perfect realism -the synthesized images are visually indistinguishable from their real-world counterparts. This area has widespread applications in industry and scientific research including 3D design, omputer animation, scientific visualization and virtual reality Facial expression synthesis has many useful applications in practice. How to synthesize facial expression images robustly and simply is still a challenging problem. This paper surveys the state of the art in this area.

show abstract

Section: Hairmentioning

confidence: 99%

Techniques of Facial Synthesis: A Comprehensive Literature Review

Chandra¹,

Karmakar²,

Hegadi³

2013

IJCA

View full text Add to dashboard Cite

show abstract

“…In computer graphics, hair acquisition, modeling, and animation have attracted growing interest in recent years [13], [14], [5], [7], [11], [22]. Hair models in graphics are threedimensional and typically have Oð10 4 À 10 5 Þ hair strands and Oð10 6 Þ line segments connected to a scalp surface.…”

Section: Introductionmentioning

confidence: 99%

“…Hair models in graphics are threedimensional and typically have Oð10 4 À 10 5 Þ hair strands and Oð10 6 Þ line segments connected to a scalp surface. These models often include hair dynamics, friction, lighting effects, and occlusion relations, which are either input through user interfaces [7], [5], [14], [13] or acquired from multiple (say 30 to 40) views using advanced computer vision techniques [17], [22]. Editing and rendering such a complex 3D model typically takes many hours on a computer.…”

Section: Introductionmentioning

confidence: 99%

A generative sketch model for human hair analysis and synthesis

Chen

Zhu

2006

IEEE Trans. Pattern Anal. Mach. Intell.

View full text Add to dashboard Cite

Abstract-In this paper, we present a generative sketch model for human hair analysis and synthesis. We treat hair images as 2D piecewise smooth vector (flow) fields and, thus, our representation is view-based in contrast to the physically-based 3D hair models in graphics. The generative model has three levels. The bottom level is the high-frequency band of the hair image. The middle level is a piecewise smooth vector field for the hair orientation, gradient strength, and growth directions. The top level is an attribute sketch graph for representing the discontinuities in the vector field. A sketch graph typically has a number of sketch curves which are divided into 11 types of directed primitives. Each primitive is a small window (say 5 Â 7 pixels) where the orientations and growth directions are defined in parametric forms, for example, hair boundaries, occluding lines between hair strands, dividing lines on top of the hair, etc. In addition to the three level representation, we model the shading effects, i.e., the low-frequency band of the hair image, by a linear superposition of some Gaussian image bases and we encode the hair color by a color map. The inference algorithm is divided into two stages: 1) We compute the undirected orientation field and sketch graph from an input image and 2) we compute the hair growth direction for the sketch curves and the orientation field using a Swendsen-Wang cut algorithm. Both steps maximize a joint Bayesian posterior probability. The generative model provides a straightforward way for synthesizing realistic hair images and stylistic drawings (rendering) from a sketch graph and a few Gaussian bases. The latter can be either inferred from a real hair image or input (edited) manually using a simple sketching interface. We test our algorithm on a large data set of hair images with diverse hair styles. Analysis, synthesis, and rendering results are reported in the experiments.

show abstract

“…Recently, researchers have shown that automatic hair geometry reconstruction based on computer vision techniques is possible [122,9]. Typically multiple images are captured about the subject head, and dense hair curve geometry is recovered from hair's photometric or stereoscopic properties.…”

Section: Related Workmentioning

confidence: 99%

“…For example, to model a digital human character, polygonal curves are often used for hair curves, and polygonal meshes for the human body surface. Although a lot of techniques and devices, such as procedural modeling [5,6], image-based modeling [7,8,9,10] and 3D scanning device [11,12,13], have been proposed to construct 3D shapes, most 3D shapes have to be modeled via user input, either from scratch or by editing / modifying existing models.…”

mentioning

confidence: 99%

Differential methods for intuitive 3D shape modeling

Fu¹

View full text Add to dashboard Cite

iv TABLE OF CONTENTS Title Page i Authorization Page ii Signature Page iii Acknowledgments iv Table of Contents v List of Figures vii List of Tables xAbstract xi In the configuration-dependent merging, the two feet are specified as handles (purple), thus remain fixed. The same boundary correspondence is used for both merging. 3.14 Our merging is applicable to a region of interest with nonzero genus. 544.1 A realistic hairstyle created using our system in five minutes. The user is allowed to design interesting hairstyles by intuitively sketching three types of style primitives: streaming curve, dividing curve and ponytail. 564.2 Left: bounding volume, vector field and style curves. Right: hair curves generated from scalp. 4.4 A dividing curve and its boundary constraints. Left: a local coordinate frame defined on the curve. Boundary constraints are defined in the local xy planes. The tilting angle α of a local plane is interpolated from two parameters α 0 and α 1 . Right: boundary constraints on a local xy plane. Blue vectors mimic the local shape of the parting hairstyle and follow the tangents of a Gaussian function, which passes through the origin p at its inflexion point and with its local shape controlled by two parameters σ and h (set to 1 and 0.5 by default, with the unit as step size in the vector field). Red vectors explicitly represent the discontinuities around the parting line in the vector field and stop hair from growing across the parting line. This thesis presents several novel intuitive techniques that make the modeling process significantly less labor-intensive. We achieve a balance between the control intuitiveness of tools and the geometric complexity of modeling output. Our techniques allow the user to intuitively control the modeling effect through only a small set of easy-to-use manipulators. The unknown vertex positions of the final models are then computed by solving a system of linear partial differential equations (i.e., the Laplace or Poisson equations) subject to the constraints derived from the manipulators. These differential-based techniques contribute to three important modeling applications: mesh deformation, mesh merging and hairstyle design. The differential mesh deformation technique allows the user to manipulate a small set of handles to deform highly detailed meshes interactively and to achieve physically plausible deformation effects. The differential mesh merging framework relieves the user's burden of both precise locating of 3D models and precise specification of merging boundaries over the meshes to be merged. For hairstyle modeling, we design a sketching interface and adapt an incremental solver to allow users to design compelling hairstyles by drawing only a small set of strokes. For all these algorithms, we pre-compute the most computationally expensive components to achieve interactive modeling with fast response.Extensive experiments demonstrate the robustness and usefulness of our techniques.

show abstract

Modeling hair from multiple views

Cited by 98 publications

References 16 publications

Techniques of Facial Synthesis: A Comprehensive Literature Review

Techniques of Facial Synthesis: A Comprehensive Literature Review

A generative sketch model for human hair analysis and synthesis

Differential methods for intuitive 3D shape modeling

Contact Info

Product

Resources

About