Conformal Equivalence

Download as pdf or txt
Download as pdf or txt
You are on page 1of 11

Conformal Equivalence of Triangle Meshes

Boris Springborn
TU Berlin

Peter Schrder
Caltech

Ulrich Pinkall
TU Berlin

Abstract
We present a new algorithm for conformal mesh parameterization. It is based on a precise notion of discrete conformal equivalence for triangle meshes which mimics the notion of conformal
equivalence for smooth surfaces. The problem of finding a flat
mesh that is discretely conformally equivalent to a given mesh
can be solved efficiently by minimizing a convex energy function,
whose Hessian turns out to be the well known cot-Laplace operator. This method can also be used to map a surface mesh to a
parameter domain which is flat except for isolated cone singularities, and we show how these can be placed automatically in order
to reduce the distortion of the parameterization. We present the
salient features of the theory and elaborate the algorithms with
a number of examples.
Keywords: Discrete Differential Geometry; conformal parameterization; conformal equivalence; discrete Riemannian metric;
cone singularities; texture mapping

Introduction

In this paper we present a definition for discrete conformal equivalence of triangle meshes and apply it to the problem of conformal mesh parameterization. Our approach arises from two basic
premises.
First, the discrete setting should parallel the smooth setting.
There the problem of finding a conformal parameterization for a
smooth surface in space is equivalent to finding a flat metric on
the surface that is conformally equivalent to the metric induced
by the embedding. Hence, we cast the parameterization problem
as one of finding conformally equivalent flat metrics for a given
metric.
Second, we treat the parameterization problem for triangle
meshes in the spirit of Discrete Differential Geometry [Bobenko
and Suris 2005]: instead of viewing discretization as a means of
making the smooth problem amenable to numerical methods, we
seek to develop on the discrete level a geometric theory that is
as rich as the analogous theory for the smooth problem. The aim
is to discretize the whole theory, not just the equations.
Instead of asking for an approximation of the smooth problem,
we are thus guided by questions like: What corresponds to a
Riemannian metric and Gaussian curvature in an analogous theory for triangle meshes? When should two triangle meshes be
considered conformally equivalent? We answer the first question
in the obvious way: the edge lengths and the angle defects at
vertices. The answer that we give to the second question is also
straightforward and intuitive. The striking point is that once we

Figure 1: Customers who visited our Tattoo parlor, where repeating


patterns can be applied seamlessly and length distortion is kept low
through cone singularities. (Insets show texture tiles.)
have thus fixed the fundamental definitions of the discrete theory,
there is an efficient algorithm to solve the discrete conformal
parameterization problem.
Contributions In this paper we present the first algorithm for

mesh parameterization which is based on a definition of discrete


conformal equivalence between triangle meshes which is satisfactory in the sense that (a) it depends only on the geometry of the
meshes and (b) defines an equivalence relation. Meshes in a conformal equivalence class are characterized by length scale factors
associated to the vertices and by conserved quantities: the length
cross ratios. This discrete notion of conformality comes with
a comprehensive theory carrying many of the hallmarks of the
smooth setting. Most importantly for graphics applications, we
present effective computational procedures based on minimizing
a convex energy. The triangulated surfaces may be of arbitrary
topology with or without boundary and no a priori cutting is
required to deal with higher genus, for example. Flexible boundary conditions support a range of options from full control over
boundary curvature to full control over length distortion, including isometrically mapped boundaries. The theory and algorithms
also admit cone singularities, which may be placed by the user

or automatically to reduce length distortion and flatten higher


genus meshes. Last but not least, our discretely conformal parameterizations admit a piecewise projective interpolation scheme
for texture coordinates, which yields improved interpolation at
no additional cost.

Conformal Equivalence

Our definitions are based on a fairly direct translation of classic


notions to the discrete setting of meshes. In smooth differential
geometry two Riemannian metrics g and g on a differentiable
2-manifold M are said to be conformally equivalent if
g = e2u g

(1)

for some smooth function u : M R, which gives the logarithm


of length change between g and g .
In the discrete setting M becomes an abstract surface triangulation consisting of vertices, edges, and triangles, M = (V, E, T ).
We do not restrict its topology or whether it possesses a boundary. While the combinatorics of a mesh are encoded in such an
abstract triangulation, its intrinsic geometry is encoded in the
edge lengths. Thus we define:
Definition 2.1. A discrete metric on M is a function l on the set
of edges E, assigning to each edge ei j a positive number l i j so
that the triangle inequalities are satisfied for all triangles t i jk T .
The smooth function u becomes a function on the set of vertices
and we define discrete conformal equivalence as:
Definition 2.2. Two discrete metrics l and l on M are (discretely)
conformally equivalent if, for some assignment of numbers ui to
the vertices vi , the metrics are related by
l i j = e(ui +u j )/2 l i j .

(2)

This notion of discrete conformality also appears in [Luo 2004].


It will turn out to be convenient to use the logarithmic lengths
i j := 2 log l i j , turning Eq. (2) into the linear relation
i j = i j + ui + u j .

(3)

Note that this notion of discrete conformal equivalence is indeed


an equivalence relation (i.e., it is reflexive, symmetric, and transitive) on the set of discrete metrics on M , and also on the set of
meshes. We consider two meshes as discretely conformally equivalent if they have the same abstract triangulation and equivalent
edge lengths according to Eq. (2).
The primary motivation for this definition is of course the obvious analogy with the smooth setting. Another is that it behaves
correctly under Mbius transformations of space: applying a
Mbius transformation to the vertices of a mesh, the resulting
mesh is discretely conformally equivalent to the untransformed
mesh. Mbius transformationscompositions of inversions in
spheresare the only (smooth) conformal transformations of
space. Thus, in particular, two meshes whose vertices are related by a smooth conformal transformation of space are also
discretely conformally equivalent. (To verify this claim, note
that the distance between two points is related to the distance
between their image points under a Mbius transformation f
by k f (x) f ( y)k = (x)( y)kx yk for a real valued function . This is obvious for translations, rotations and scaling,
and a straightforward calculation shows that it is true also for
inversions in a sphere.) As a practical consequence, this Mbius
invariance of discrete conformality enables us to map meshes not
only to the plane but also to the sphere.
Discrete conformal equivalence can also be characterized in terms
of conserved quantities, namely the length cross ratios:

Definition 2.3. Given a discrete metric l, we associate with each


interior edge ei j (between t i jk and t jim ) the length cross ratio

ci j := l im /l m j l jk /l ki .

(4)

Proposition. Two meshes are discretely conformally equivalent if


and only if their length cross ratios are the same.
Proof. In one direction, the implication is obvious: if the meshes
are discretely conformally equivalent, then their length cross
ratios are equal because the scale factors eu/2 cancel. To see the
converse, consider two discrete metrics l and l. We have to show
that if their length cross ratios c and c are equal, then we can
associate numbers ui to the vertices vi satisfying Eqs. (2) (one
equation for each edge). In general, this system of equations
for the ui is overdetermined. For each triangle t i jk the three
equations corresponding to its edges already determine values
for u at its vertices:
eui = l i j /l i j l ki /l ki l jk /l jk .

(5)

Considering the neighboring triangle t jim one obtains the same


value for eui precisely if ci j = ci j . Hence, if l and l have identical
length cross ratios, values for u at the vertices are consistently
determined by Eqs. (2).
The discrete theory we have set up now informs our consequent
approach. Given a mesh M and a discrete metric l on it, we
consider the equivalence class of conformally equivalent metrics.
In this class we look for a metric l which is flat. Only in a
subsequent step (Section 3.3) do we construct vertex coordinates
which realize this flat metric. To find the desired flat metric we
need the relation between length and curvature.

2.1

From Curvatures to Lengths

Given a triangle t i jk T and lengths l i j , l jk , l ki satisfying the


triangle inequality, the angle ijk at vertex vi opposite edge e jk
can be recovered by using the cosine formula or the half-angle
formula:

ijk = 2 tan1

(l i j +l jk l ki )(l jk +l ki l i j )
(l ki +l i j l jk )(l jk +l ki +l i j )

(6)

as a function of l). Given the angles, curva(and similarly for


tures follow as the excess angle sums at vertices. The angle sum
at a vertex vi is denoted by i :
P
i = t i jk 3vi ijk
e
(and similarly for ).
The curvature at an interior vertex is
Ki = 2 i and the boundary curvature at a boundary vertex
e,
).
is i = i (and similarly for K
Our goal now is to find a new metric, which is conformally equivalent to a given one, and which possesses prescribed curvatures.
Thus, we have to solve the following problem:
b i , find ui
Problem. Given, for each vertex, a desired angle sum
e i =
b i.
such that the new metric l has angle sums
e i =0,
e i =2 at interior vertices, i.e., K
For instance, if we prescribe
a solution to our problem will give a conformally equivalent flat
i.
metric with prescribed boundary curvatures
The following conditions on the prescribed angle sums are
necessary for the existence of a solution. They must satisfy
b < |{t i jk 3 vi }| because the triangle angles are between
0<
0 and , and they must be compatible with the Gauss-Bonnet
formula
P
P
vi M \ M Ki +
vi M i = 2(M ),
where (M ) = |T | |E| + |V | is the Euler characteristic of M .

In view of Eqs. (2) and (6), the problem is equivalent to solvei


ing a set of n := |V | equations in n variables ui . Since the
automatically satisfy one linear relation (Gauss-Bonnet) and the
angles are scale invariant, we really have n 1 conditions for
n 1 essential
Pdegrees of freedom. We fix the scaling ambiguity
by requiring vi ui = 0.
Luckily, if a solution to this set of non-linear equations exists, it
can be found as the unique minimizer of a convex energy in u
X
X

ij,
jk ,
ki ) (ui +u j +uk ) + 1
b i ui (7)
E(u) =
f (

2
2
t i jk T

vi V

where
ij,
jk ,
ki ) =
f (


j
j
1
jk +
i j + L(
ijk
ki
kij
ki ) + L(
ijk ) + L(
kij ),

ki +
2

(8)

as in Eq. (3); L(.) denoting Milnors Lobachevsky function


with
and
depending on u.
(see Appendix A); and both
Indeed, the partial derivatives of this energy are
P

bi
ijk ,
ui E = 12
t i jk 3vi

Circle Patterns Going back to an unpublished idea of Thurston


(see [Stephenson 2003] for an eyewitness account), discrete
conformality can be approached through circle patterns. Here
meshes are considered together with a system of circles associated to faces [Kharevych et al. 2006] or vertices [Bowers and
Hurdal 2003; Stephenson 2005; Jin et al. 2007; Yang et al. 2008].
Two such meshes are considered conformally equivalent, if the
intersection angles of the circlesor inversive distances for nonintersecting circlesare equal in both meshes.

These methods require the solution of non-linear equations, using a convex variational principle, a specially tailored relaxation
procedure, or a flow.
This approach to discrete conformality is also not fully satisfactory.
In the case of Kharevych et al., the intersection angles determine
the curvature of the mesh. Consequently, they must change when
the mesh is flattened. For the other circle methods, the notion
of discrete conformality either depends on an arbitrary choice
of circles, or it is defined only if the edge-lengths satisfy some
further conditions, or both.
Curvature Flow Consider u as a function of time and evolving

(9)

so that grad E(u) = 0 iff u solves the problem. The Hessian of the
energy turns out to be one half of the well-known cot-Laplace
operator:
P
(Hess E u)i = 12 (u)i = 14 ei j 3vi w i j (ui u j ),
(10)
kij + cot
li j for interior edges and only one cot
with w i j = cot
term for boundary edges. (We adopt the sign convention for
the Laplace operator which renders it positive semi-definite. See
Appendix B and C for proofs of Eqs. (9) and (10).)
The Hessian is therefore positive semi-definite with only the constant functions in its null-space. This corresponds to the fact that
the energy E(u) is scale invariant, i.e., its value does not change
if the same number is added to all ui .

under the negative gradient of E


bK
e) .
t u = 2 grad E = (K
(For simplicity, assume that M has no boundary.) We may then
think of u = argminu E(u)which we compute using Newtons
methodas the steady state solution of this curvature flow with
e = u K
e t u together with
given target curvatures. Using t K
Eq. (10) we also find that the evolution of the curvature is governed by the equation
e = (
bK
e)
e K
t K
e denoting the (positive semi-definite) cot-Laplace operator
with
of the discrete metric induced by u(t).

There are many algorithms for mesh parameterization in the


literature and we will not attempt a comprehensive review here
(the interested reader is referred to [Floater and Hormann 2005]
and [Sheffer et al. 2006]). Instead we focus on approaches which
are based on discretized or discrete notions of harmonicity and
conformality.

b = 0, this flow was considered by Luo [2004]


In the case where K
as a discrete version of Yamabe flow. He proves that the flow
is variational, but he gives no formula for the energy and does
not seem to be aware that the cot-Laplace operator appears in
e . A different discrete curvature flow
the evolution equation for K
forms the foundation of the approach of Jin et al. [2007] who
use Chow and Luos definition of discrete Ricci flow [2003]. (For
2-dimensional Riemannian manifolds, Yamabe flow is the same
as Ricci flow.) Our explicitly variational formulation provides a
new approach to discrete Ricci flow for surfaces.

Discrete Harmonicity Discrete versions of the theory of harmonic

Angle Based Flattening Smooth conformal maps preserve an-

2.2

Relation to Other Approaches

and analytic functions were developed as early as the 1940s and


50s [Duffin 1956], based on simple difference equations analogous to Laplaces equation and the Cauchy-Riemann equations.
Indeed, the discovery of the cot-Laplace operator heralds from
this time [Duffin 1959], and these ideas still inform contemporary notions of discrete conformality and harmonicity that are
based on linear conditions on the vertex coordinates. Examples
of theory and applications include [Pinkall and Polthier 1993;
Mercat 2001; Desbrun et al. 2002; Lvy et al. 2002; Gu and Yau
2003; Tong et al. 2007].

gles exactly. Enforcing this property as-best-as-possible in the


discrete setting forms the foundation for Angle Based Flattening
(ABF) [Sheffer et al. 2005]. The resulting optimization problem
for the angles of the flat mesh is non-linear with non-convex
non-linear constraints. It requires sophisticated numerical methods [Sheffer et al. 2005] (for some recent progress see [Zayer
et al. 2007]), and is not solved to high enough accuracy to actually produce a flat metric. Thus a further approximation step is
necessary to produce the final coordinate functions.

The linear theory of discrete harmonicity is interesting and rich,


attractive computationally, and enormously useful in applications.
But the implied notion of a discretely conformal map as a pair of
conjugate discretely harmonic functions is deficient because the
inverse of such a map, and the composition of two such maps, are
no longer discretely conformal. Thus, this notion of a discretely
conformal map does not define a notion of discrete conformal
equivalence.

Metric Scaling There is one very recent approach that is closely

related to ours [Ben-Chen et al. 2008]. Just as we do, they first


compute a metric for the image mesh and only then a set of
vertex positions. They start with the following well known relation. Given a smooth 2-manifold and two metrics g, g related
through the conformal factor e2u as in Eq. (1)the curvatures K,
e satisfy:
K
e = K + u,
e2u K
(11)

where is thepositive semi-definiteLaplace-Beltrami operae = 0), if u solves the Poisson equation


tor. The metric g is flat (K
u = K.

(12)

To flatten a mesh, Ben-Chen et al. proceed by solving the discretized Poisson equation with the cot-Laplace operator in the
original metric and Ki the angle defect at a vertex. This is precisely the first Newton step when minimizing our energy E(u).
(Their definition of length change induced by the ui is different
from Eq. (2) but this is immaterial to the present argument.) As
Ben-Chen et al. point out, this does not in general yield a flat
metric. They address this by solving a cot-Laplace layout problem
in the new metric to find flat vertex positions which approximate
the desired metric. In contrast we solve for a (numerically) flat
metric and require no further approximation to recover the actual
vertex positions.
In their method, target curvatures at the boundary as well as for
cone singularities are computed based on a diffusion problem and
its associated Greens functionsone each per boundary vertex
and interior cone vertex. Incidentally, this computational burden
could be removed by using our natural boundary conditions and
free cone angles (Sections 4.2 and 5.1).
2.2.1

Discussion

In contrast to all other approaches, ours is based on a proper


notion of discrete conformal equivalence for triangle meshes.
This notion depends on the meshes alone without any arbitrary
choices, and it is adhered to not approximately but precisely.
We have an explicit expression for the energyas do Kharevych
et al. but not Jin et al.which facilitates the use of standard,
globally convergent, Newton trust region methods. We can also
accommodate cone singularitiesas introduced by Kharevych
et al. and recently demonstrated for discrete harmonic approaches [Tong et al. 2007; Klberer et al. 2007]and do not
require a priori cutting for higher genus surfacesas is done
in ABF. Boundary conditions include constraints on angles or
lengths andnovel among all approacheswe support isometrically mapped boundaries. Incidentally, the latter also pick out
the unique least distorted mapping in a given equivalence class
(more on these aspects in Section 4).
Comparing our method with ABF illustrates how choosing the
right discrete notions can lead to concrete practical benefits.
ABF preserves triangle angles as much as possible. Our approach
preserves the discrete conformal structure of the mesh precisely.
As it happens, this leads to a mathematically simpler optimization
problem.
After this discussion of related approaches in computer graphics, we want to acknowledge at least briefly some mathematical
work we have built upon. Troyanov [1986] treated the problem
of finding a conformally equivalent flat metric with prescribed
cone singularities in the smooth setting and proved existence and
uniqueness of the solution. The construction of our energy relied
heavily on previous work on variational principles for circle patterns [Colin de Verdire 1991; Rivin 1994; Leibon 2002; Bobenko
and Springborn 2004]. There is a fruitful connection between
all of these circle pattern energies and our energy E(u) on the
one side and hyperbolic geometry on the other side. A reader familiar with hyperbolic geometry will realize that the appearance
of Milnors Lobachevsky function is a hint that our energy E(u)
has something to do with the volumes of ideal hyperbolic tetrahedra, and she may recognize the logarithmic edge lengths i j and
length cross ratios ci j as Penner coordinates and Thurston-Fock
shear coordinates for the Teichmller spaces T g,n of genus g Riemann surfaces with n punctures. In fact, minimizing E(u) also

solves the problem of finding polyhedral realizations for complete hyperbolic surfaces. But this side of the story is beyond the
scope of the present paper.

Discretely Conformal Parameterization

Now that the theoretical foundations are laid down we proceed with a discussion of our basic conformal parameterization
method.

3.1

Convex Optimization

So far we have a convex energy, but we do not yet have a convex


optimization problem since the domain, i.e., the set of u resulting
in new edge lengths satisfying the triangle inequalities, is not
convex. The inset figure below illustrates the legal range (yellow)
of u values for an example triangle. We get around this difficulty
by extending the domain of E(u) to all of Rn :
if l jk
l ki + l i j

then

ijk = and
ki =
kij = 0.

Now Eqs. (7) and (8) define E(u) for all


values of u. This simple way of extending
E(u) is C 1 . The gradient of the extended
energy is still given by Eq. (9), and the
Hessian is given by Eq. (10), where, how kij in the equation for
ever, the terms cot
the weights must be replaced by 0 when kij = 0 or . The figure on the right
ever
indicates the iso-contours of the extended
energy.

u3
u2
u1 +u2 +u3 =0
12 =23 =31 =0

We now have a standard unconstrained convex optimization problem with explicit formul for the target function, its gradient,
and Hessian. The Hessian is positive semi-definite with only the
constant vector in its null-space, reflecting the fact that E(u) remains invariant if a constant vector is added to u. To find the
argmin of the extended energy we use the globally convergent
Newton-Steihaug trust region method [Steihaug 1983] as implemented in PETSc/TAO [Balay et al. 2007; Benson et al. 2007].

3.2

Violation of the Triangle Inequality

When minimizing the extended energy it is possible that the


global minimum is achieved for a u giving l which violate the triangle inequality. Figure 2 shows a sampling of such cases which
are representative of our experience. In all the cases we have
observed we found that the triangulation near the bad edge was
highly degenerate. In almost all cases this was due to a triangle
with one angle close to . In others, multiple triangles were
folded over one another (see the second and third example in
the top row and first example in the bottom row of Figure 2).
Unless the mesh has regions full of such degenerate situations,
the problem is simply fixed by flipping or alternatively subdividing the edges opposite the straight angles. Our worst example
was the Gargoyle which required flips for 55 out of 74964 edges.
(Edge flipping was also considered by Luo [2004], albeit during
his curvature flow.)

3.3

Layout

With u = argminu E(u) we have new lengths l and angles ,


but not yet a new mesh. For texture mapping we need vertex
positions which requires a layout procedure. Suppose M is a
topological disk and there are no cone singularities in the interior.
We traverse the dual graph of M in a breadth first manner, laying
out opposing vertices each time an edge is crossed. The orientation of each edge is determined by summation of corner angles
along the traversal. With this layout procedure meshes with 100s
of thousands of vertices and length ratios as high as 106 can be
handled with high precision. Typical examples start with metric

geneous coordinates. The figure above shows a comparison between linear interpolation (top) and projective interpolation (bottom) with some notable differences highlighted.

Boundary Conditions

Both as a practical matter and as a key to achieving functionality relevant in applications, boundary conditions of different
types can be employed. We review these in turn, using practical
examples to motivate them.

4.1
Figure 2: A sampling of ripped edges. In all cases the local
geometry was near singular and edge flipping (alternatively, edge
subdivision) remedied the trouble.
data flat to within k grad E(u )k2 < = 1014 . . . 1012 and yield
worst relative length errors bounded by 10 to 1000.
Arbitrary topology meshes and those with cone singularities are
reduced to the topological disk case. To resolve handles we compute a system of loops [Erickson and Whittlesey 2005]. Cutting
the loops gives a topological disk, possibly with cones in its interior. To resolve the cones, we trace a path from each cone back to
the base vertex of the system of loops using the Dijkstra tree from
the first phase. (Any boundaries, including the one due to the
system of loops, are treated as if each were a single cone vertex
from the point of view of spanning tree construction.) Cutting
this cone spanning tree yields a topological disk with no cones
in the interior and it may be laid out as above. We have made
no efforts to move the paths to inconspicuous locations, though
methods such as Seamster [Sheffer and Hart 2002] could be employed. Note that vertices along the cut path will have multiple
positions in the layout.

3.4

Piecewise Projective Interpolation

After vertex positions for the image mesh


have been found, it is usually necessary,
e.g., in texture mapping, to extend the
map from the vertices to the whole mesh
surface through interpolation. Typically
this is done piecewise linearly. For conformally equivalent meshes there is another possibility unique to them.
Given a domain and range triangle,
there is always a unique projective transformation mapping one onto the other
and the circumcircle of one onto the circumcircle of the other. In general, these
projective mappings do not fit together
continuously across edges. A unique exception are the discretely conformally
equivalent meshes: This circumcircle
preserving piecewise projective interpolation is continuous across
edges iff the meshes are discretely conformally equivalent. This
follows from the proposition in Appendix D, which also shows
that in practice it is rather easy to take advantage of this (and we
do so in all our texture visualizations): If (x i , yi , zi ) and (si , t i )
are the vertex coordinates for the original and image mesh respectively, the correct projective interpolation is achieved by performing linear interpolation on the homogeneous coordinates
(x i , yi , zi , 1) and eui (si , t i , 1). (Note the scale factor eui which
is applied to the homogeneous coordinates of the image mesh.)
Since conventional graphics cards support linear interpolation
between homogeneous coordinates for texturing anyway, the implementation of the piecewise projective interpolation scheme
involves nothing more than using these properly scaled homo-

Fixed Boundary Curvature

The Problem posed in Section 2.1 requires us to prescribe ane i at inteb i for all vertices, i.e., target curvatures K
gle sums
rior verticesmostly 0 unless we want a cone singularityand
i for boundary vertices. For example, to
boundary curvatures
achieve a rectangular layout, which has obvious advantages for
e i = for all boundary vertices but
texture packing, one sets
e i = /2 (Figure 3, left; see also the Stamp
four, which receive
part in Figure 14).

Figure 3: Fixing curvatures on the boundary can be useful for


achieving a particular boundary shape. In contrast, natural boundaries do not constrain the boundary curvature at all (right) and
yield the least distorted mapping.

4.2

Free Boundary Curvature

b at some vertices
Alternatively, we may prescribe angle sums
and fix u at others. In this case the unique solution can be
found by considering the fixed ui as constants and minimizing
E(u) as a function of the remaining variables. Since the value
b at vertices with fixed u is clearly irrelevant for the
assigned to
minimization we may leave it undefined at these vertices. With
some ui fixed, the problem is not scale-invariant anymore and
the reduced Hessian becomes positive definite.
An extreme example is to fix u for all boundary vertices. This
gives complete control over the length distortion along the boundary while leaving the boundary curvature free.
Natural Boundary Conditions One choice is to set ui = 0 for

all boundary vertices. In that case all boundary edges retain


their original length, i.e., the boundary is mapped isometrically.
e i = 2 for all
Figure 3 (right) shows a mesh flattened with
interior vertices and ui = 0 for all boundary vertices. We call
these boundary conditions natural because of the following
remarkable fact from the smooth theory: Among all flat metrics
g that are conformally equivalent to a given metric g, one with
least distortion is obtained by setting u = 0 on the boundary
(where the distortion is measured by the Dirichlet energy of u;
see Appendix E).

4.3

Topological Spheres

Meshes with sphere topology are sufficiently common that specific parameterization algorithms for them have been designed.

Like Kharevych et al. [2006], we exploit the Mbius invariance


of discrete conformality to map a mesh to the unit sphere. First
some vertex is sent to infinity through inversion. Removing its
b i = 2
vertex star we fix ui = 0 for the vertices of its link, and
for all other vertices. The resulting planar layout is projected
stereographically to the sphere and the knocked out vertex reinserted at the north pole (Figure 4). Subsequent Mbius normalizations can be applied to ensure, for example, that the center of
mass of all vertices is at the origin [Springborn 2005]a notion
of equidistribution of points. A simple calculation shows that
the resulting mesh is discretely conformally equivalent, even at
the knocked out vertex.

with cone angles restricted to positive integer multiples of /2


were used in rectangular surface patching [Ray et al. 2006] and
in discrete harmonic methods [Tong et al. 2007; Klberer et al.
2007] for the construction of union-of-quad parameterizations.
Aiming to align iso-parameter lines (roughly) with principal curvature directions, these approaches all placed cone singularities
at numerically estimated umbilic points.
Our method supports free and fixed cone angles (Figure 5) and
we are interested in placing cone singularities in a manner which
reduces length distortion, i.e., reduces the variation of u, and to
do so automatically.

5.1 Automatic Cone Singularities

Figure 4: Sphere parameterizations of Dragon and Hygeia visualized as spheres with the original normals.

Cone Singularities

In practice, a significant issue with conformal maps is their at


times severe length distortion. Gu et al. [2002], for example,
added cuts into the worst distortion area, repeating as necessary.
To avoid cutting and the associated discontinuities or complex
boundary conditions, Gu and Yau [2003] used topological puncturing and double cover constructions.
The most general and flexible approach to date though was introduced by Kharevych et al. [2006] and is based on the selective
introduction of isolated vertices which are not required to have
zero Gaussian curvature. At these cone vertices, the local metric
ei =
is that of a Euclidean cone with some cone angle
6 2. This
was used both to accommodate higher genus surfaces and reduce
length distortion. Placement relied exclusively on human geometric intuition and trial and error. More recently cone singularities

Where should the zero curvature assumption be relaxed so as to


reduce distortion? Considering Eq. (12) for the smooth setting,
we see that a local change in K leads to a localized change in
u, since adding a Dirac to the right hand side has the effect of
adding a Greens function to the solution u. Thus, we choose
the vertex with the worst distortion, vi = argmax vi |ui |, as the
location for a cone singularity. (If a boundary is present
we
P
assume natural boundary conditions, otherwise we set ui = 0.)
Instead of prescribing a cone angle, we leave it free, setting ui = 0
(Section 4.2). Since a cone singularity may be seen as the limiting
case of a very small hole, this makes sense in view of the minimal
distortion property of natural boundary conditions. Our strategy
for the placement of the cone singularities is thus very similar to
the one of Ben-Chen et al. [2008], but we do not have to do any
extra work to determine the cone angles.
A possible greedy approach starts with some minimizer u of E
and selects the vertex with the largest magnitude ui , setting it to
free cone status, and repeating. Figure 6 shows a sequence of
such greedy free placements and their impact on u.

max |u| = 1.84

max |u| = 1.16

max |u| = 0.86

max |u| = 0.67

Figure 6: Visualization of u as a graph over the domain for the


mannequin head. Using natural boundaries, zero to three greedy
free cones are placed, significantly reducing the magnitude of u in
each step.

Figure 5: The effect of cone singularities. Beginning with natural


boundary conditions (left), two free cones are added automatically
(middle) to reduce distortion (Section 5.1). Depending on the
texture, seams may appear. Choosing a suitable cone angle (here
) can remove this problem (right). All three flat metrics are in the
same equivalence class as the original metric.

If the mesh is a closed surface of genus 6= 1 we cannotdue to


the Gauss-Bonnet relationget an initial solution for u without
prescribing initial curvatures at vertices. Instead of concentrating
all of the required total curvature at some vertex like Ben-Chen et
al., we distribute the required total curvature over the whole
mesh such that each vertex receives an equal share initially. This
is done for the initial solve only. As free cones are placed, the
remaining vertices revert to the desired zero curvature setting.
Because the greedy procedure cannot undo earlier decisions, it
comes as no surprise that its results can be improved (Figure 7).
Four free cone singularities were placed in a greedy manner

Figure 7: Comparison of the greedy strategy (left) with concurrent


cone optimization (right) for four cones.
resulting in two cones on top of the head (front and back), at the
left ear and on the nose (left). Starting from this initial placement
of the four cones, several sweepsthree in this caseof a nonlinear Gauss-Seidel solver result in the placement on the right.
Cones are now placed far more symmetrically: at the top of the
head, on each ear and on the nose. Typically it takes two or three
rounds for the cone positions to settle. To save effort during
the Gauss-Seidel stage, we do not solve for u to high accuracy
(only one Newton step is used). The final u-function is solved
to high precision. Figure 8 shows another where this procedure
was applied with a total of 18 cones.

Figure 8: Oliviers hand with 18 optimized free cones (distributed


over the finger tips and inbetween the fingers).

Texture Rectification

When mapping unstructured textures, minimizing distortion in


the parameterization is an important criterion. Using textures
with symmetries, e.g., checker boards, or when using parameterizations for purposes of resampling, additional constraints on
cone placement and cone angles are called for. Quadrilateral
b i = k/2, 0 < k 6= 4.
patching in particular calls for cones with
A simple example of this is seen in Figure 5, where two free cones
were quantized to a fixed angle of each. Layout followed by
a suitable rigid motion and global scale places the two cones
at (0, 0) and (1, 0) in the parameter plane. A seamless checker
board texture results.

Figure 9: A layout for the Venus with four cones of each. These
map to corners of a parallelogram (left side). Using a checker board
texture leads to artifacts (inset). These can be removed by rectifying
the layout with an additional transformation (right side).
values may violate the Gauss-Bonnet relation, for example, the
sum of quantized cone values may be too large. In that case the
cone angle that was rounded up the most has its quantized value
decreased (unless it is /2 already). Repeating this procedure if
necessary leads to a set of quantized values which minimize the
maximum error between orginal and quantized values while satisfying the Gauss-Bonnet relation. (An analogous procedure can
be applied if the initial sum is too small.) Using these now fixed
cones the corresponding conformally equivalent metric is solved
for and laid out (Section 3.3). It defines the closed meta-polygon
P = {pi C | i = 0, . . . , m 1} given by the vertex positions
treated as complex numbersof m special vertices encountered
in a single, ordered traversal of the layout boundary. Special vertices are (1) cone vertices, (2) interior vertices with more than
two cut edges incident on them, and (3) boundary vertices with
a cut edge incident on them.
Figure 10 shows a meta-polygon for the Lion dataset. Five free
cones were placed with the simultaneous optimization procedure, quantized to the nearest multiple of /2 (3/2 for all five
cones in this case), and the corresponding conformally equivalent
metric laid out.

But quantizing the cone angles to integer multiples of /2 is


generally not enough if there are more than two cone vertices.
Figure 9 illustrates such an example. Four cones of each were
placed on the Venus mesha topological sphere. The resulting
layoutregardless of the particulars of the cone spanning tree
must arrange the cone images as the corners of a parallogram,
but generally not as the corners of a rectangle with integer side
lengths. In the case of the Venus example a simple shear and
scale is enough to produce a seamless checker board texture.
(Afterward the metric is no longer conformally equivalent.)
Let us consider this problem in greater generality. Assume M
is a topological disk or sphere (the reason for this restriction
will be explained below) and any cones multiples of /2. For a
topological disk, any free conesresulting from our automatic
procedure (Section 5.1)are rounded to the nearest positive
multiple of /2. In the case of a closed surface, the quantized

Figure 10: Meta-polygon for the Lion with quantized cones (solid
red) before rectification. After rectification, which forces all cone
points to integer locations, the texture tiles the model seamlessly
(inset shows 3/2 cone on chin underside).

Meta-edges correspond to paths of edges along the boundary of


the layout, joining two consecutive special vertices. Such a path is
either along the boundary of the original mesh (black edges), or
one side of a cut (colored edges), with the other side (matching
color) corresponding to a different meta-edge. The two sides
of a cut path fit together: a Euclidean motion moves one to the
other, and the same motion aligns the corresponding meta-edges.
As Figure 10 suggests, the rotation angles of these motions are
multiples of /2. This is true in general due to the quantized
cone angles, but only if the original mesh is simply connected
hence the topological restriction to disks and spheres.
Suppose for a moment that the special vertices of type (1), the
cone vertices, have integer coordinates. Then a checkerboard
pattern with these vertices in the centers of the squares would fit
seamlessly on the mesh, because the Euclidean motions aligning
corresponding meta-edges would be symmetries of the pattern.
In reality, the coordinates are not integer. (Figure 11 shows an
example with eight optimized and quantized cones, which are
not on integers; note the apparent rip in the texture on the far
end of the right wing.) How close they are to being integer, and
hence how much distortion is needed to make them so, depends
on rotation and scaling: the larger the scale, the finer the integer
grid is relative to the layout size. Suppose such a rotated and
scaled layout is given. The idea now is to deform the metapolygon so that all cone vertices become complex integers, Z2 :=
Z + IZ C, and adapt the layout accordingly. The result will no
longer be exactly discretely conformal, but only approximately
so: Seamless alignment can only be achieved by relaxing the
conformality requirement.

Figure 11: The Gargoylea topological spherewith eight optimized and subsequently quantized cones. On the top right the
optimized cone domain and in the center the result of quantizing
the cones. Since the cones do not have integer locations the checkerboard texture has seams.

6.1

of (column) basis vectors spanning the null space of A, chosen so


that all entries of NA are complex integers.
Now assume the vertices of the meta-polygon are numbered so
that p0 is a cone vertex and translate it toPthe origin. Then all
k1
cone vertices have integer coordinates if i=0 di Z2 for all k
such that pk is a cone vertex. This can be written as
B d (Z2 )l
where B is a matrix with m columns and l rows (one less than
there are cone vertices). Finally, the columns of BNAall entries
are complex integersspan a complex subspace which has nonempty intersection with the lattice (Z2 )l . From this subset lattice
we wish to select the nearest (in some norm) lattice point to
the cone positions in the given (non-integer) layout. This is
an instance of the classic closest lattice vector (CLV) problem.
Optimal solutions to this problem are computationally hard. For
the examples in this paper we proceeded by computing the row
rank of BNA, fixed as many (l 1) of the cones to the nearest
point in Z2 and solved for the remaining (l th ) cone location.
Given a solution d the implied deformation must now be interpolated across the layout. We do this by performing a cot-Laplace
layout as explained in the next section.

6.2

Generalized Laplace Layout

In Section 3.3 we described our breadth first layout procedure to


turn metric data into coordinate functions. Another way to turn
metric data into coordinate functions is through the use of the
cot-Laplace operator.
For a flat metric the cot-Laplace operatoras a map Cn Cn .
(n = |V |) and using natural boundary conditions [Desbrun et al.
2002; Lvy et al. 2002]has a two dimensional (complex) null
space consisting of the constant and linear functions on the vertices. The layout procedure produces the uniqueup to scale and
a rigid motionnull space vector orthogonal to the constant vector. We now turn this prescription around: Given a cot-Laplace
operator due to a flat metric, produce the null space vector with
zero mean. Because of the presence of cone singularities this
cot-Laplace operator must incorporate the edge identifications
via rigid motions encoded in the meta-polygon. This is done in
the same manner as [Tong et al. 2007]. (Their method accommodates general rotations with no problem including the fact
that only a single complex variable is needed for each vertex.)
Cone singularities enter as Dirichlet data with positions. The
cot-Laplace operator now has only one remaining vector in its
null space: the original layout (this is true even for general cone
angles).

Deforming the Meta-Polygon

When deforming the meta-polygon, associated pairs of metaedges must maintain their alignment through Euclidean motions
with rotations being multiples of /2. This requirement is best
treated by considering the meta-edge differences pi+1 pi =: di
C (all index arithmetic is understood modulo m). Meta-edges
due to cutting come in pairs (i, j) and we have
di = R i j d j

with R i j {+1, +I, 1, I}.

The perturbed edge differences


Pd must maintain these relations
and keep the polygon closed:
di = 0. This can be written as
Ad = 0,
where A is a matrix with m columns and one more row than
there are pairs of corresponding meta-edges. Let NA be a matrix

Figure 12: Rectified domain for Max Planck dataset and visualization of corresponding iso-lines on the mesh.
We have run comparisons between breadth first and Laplace
layout, and found the results to be of comparable accuracy even

for very large layouts. Solving large cot-Laplace systems to high


accuracy is in general computationally far more demanding than
a simple breadth first traversal, of course.
If the meta-polygon is deformed and we use the new cone locations as Dirichlet data the cot-Laplace operator will generally
not have a null space anymore and we return the vector of least
Dirichlet energy. Figure 12 shows the rectified domain and corresponding iso-lines on the model of Max Planck; see also Figure 10
for the rectified Lion head, and Figure 13 for the rectified Torso
parameterization.

very degenerate in many places and the trust region solver took
very small steps. As a quantitative distortion measure we used
the area weighted mean of quasi-conformality [Kharevych et al.
2006] (denoted kqck1 in the table), i.e., the ratio of larger to
smaller singular value of the surface to parameter plane tangent
map. A value of one indicates no angle distortion at all. While it
is difficult to compare these numbers across papers ours appear
generally somewhat worse than what was reported by Ben-Chen
et al. and Kharevych et al., for example. Given that we effectively
enforce more structure than Ben-Chen and co-workers, and do
not allow any approximation of the original data as Kharevych
and co-workers do, this is perhaps not surprising. The last column
in the table indicates the boundary conditions (free, rectangle,
sphere) or the number of optimized (o) free cones, whether they
were subsequently quantized (q), and/or rectified (r).
Model
|V | Iter. kukmax
Face
1042 16 1.164
Stamp
4100
6 0.431
Stamp
4100
6 0.742
Lion
8356
6 1.089
Olivier
24795
6 1.188
Max Planck 25445 11 0.908
Gargoyle
24990 56 1.849
Hygeia
140654
5 0.650
Torso
142348 12 1.957
Torso
142348 21 1.091
Dragon
152803
5 3.759
Julius
209083
5 1.173

Figure 13: Visualization of the parameterization on the Torso with


14 optimized, quantized, and subsequently rectified cones (each ear,
each nostril, either side of back of neck, on Adams apple, three on
each arm stump and one cone on the torso bottom.

Numerical Results

We have applied our techniques to a number of meshes and report


a representative subset of the results here. Mesh sizes ranged
from a few hundred vertices to over 200,000 (Julius dataset in
Figure 14). In all cases we achieved gradient magnitude residuals
in the 1014 . . . 1012 range. None of our layoutseither breadth
first or the cot-Laplace operator for rectified layoutshave any
flipped triangles. The maximum, relative length layout error
was always bounded by 10 to 1000 times the gradient residual
norm. The Newton trust region solver typically requires around
10 iterations, each iteration having a cost proportional to the
assembly and solve phases of a single cot-Laplace problem. The
number of Newton steps is thus the relevant figure of merit in
terms of overall runtime. (Absolute performance numbers are
meaningless since individual solver implementations and processor differences can easily account for factors of 10 and more
in observed wall clock times. With this proviso, we note that
PETSc/TAO running on a 1.1GHz Pentium M requires 162s for
Julius and 230s for the Torso.) The case of the Gargoyle is exceedingly atypical in our experience. This particular mesh is

kqck1
cones
1.0702 rectangle
1.0541
free
1.0625 rectangle
1.0282
5 (oqr)
1.0087
18 (o)
1.0097
4 (oqr)
1.0509
8 (oq)
1.0120
sphere
1.0048
4 (o)
1.0036 14 (oqr)
1.0277
sphere
1.0027
free

Conclusion

We have shown that our simple notion of discrete conformal


equivalence leads to attractive algorithms for mesh flattening,
including the automatic determination of cone singularities to
reduce distortion and deal with higher genus surfaces, as well as
a new piecewise projective interpolation scheme. In Section 6
we have taken first steps to tackle the problem of mapping a symmetric pattern seamlessly onto an arbitrary surface. We have discussed a way to deal with cone singularities, but how to achieve
texture rectification for surfaces of higher genus remains an open
question. Here we see a promising new direction for further
research.
Acknowledgment This work was supported in part by NSF (CCF-0528101 and CCF0635112), DOE (W-7405-ENG-48/B341492), the Caltech Center for Mathematics
of Information, DFG Research Center Matheon, the Alexander von Humboldt
Stiftung, and Autodesk. The authors are gratefully indebted to Alexander Bobenko
for inspiring discussions. Special thanks to Cici Koenig, Andreas Fabri, Pierre Alliez,
and Mathieu Desbrun.

References
Balay, S., Buschelman, K., Eijkhout, V., Gropp, W. D., Kaushik,
D., Knepley, M. G., McInnes, L. C., Smith, B. F., and Zhang, H.
2007. PETSc Users Manual. Tech. Rep. ANL-95/11 (Revision
2.3.3), Argonne National Laboratory. http://www.mcs.anl.
gov/petsc/.
Ben-Chen, M., Gotsman, C., and Bunin, G. 2008. Conformal
Flattening by Curvature Prescription and Metric Scaling. Comp.
Graph. Forum 27, 2, 449458.
Benson, S., McInnes, L. C., Mor, J., Munson, T., and Sarich,
J. 2007. TAO User Manual. Tech. Rep. ANL/MCS-TM-242
(Revision 1.9), Argonne National Laboratory. http://www.mcs.
anl.gov/tao.
Figure 14: Further examples of a free boundary (Julius) and use of
boundary curvatures to enforce a rectangular domain (Stamp).

Bobenko, A. I., and Springborn, B. A. 2004. Variational Principles


for Circle Patterns and Koebes Theorem. Trans. Amer. Math.
Soc. 356, 2, 659689.

Bobenko, A. I., and Suris, Y. B., 2005. Discrete Differential


Geometry. Consistency as Integrability. Preprint arXiv:math/
0504358v1. To appear in Graduate Studies in Mathematics of
the AMS.
Bowers, P. L., and Hurdal, M. K. 2003. Planar Conformal Mappings of Piecewise Flat Surfaces. In Vis. and Math. III. Springer,
334.
Chow, B., and Luo, F. 2003. Combinatorial Ricci Flows on Surfaces. J. Diff. Geom. 63, 1, 97129.

Rivin, I. 1994. Euclidean Structures on Simplicial Surfaces and


Hyperbolic Volume. Ann. of Math. (2) 139, 553580.
Sheffer, A., and Hart, J. C. 2002. Seamster: Inconspicuous
Low-Distortion Texture Seam Layout. In Proc. IEEE Vis., IEEE
Comp. Soc., 291298.
Sheffer, A., Lvy, B., Mogilnitsky, M., and Bogomyakov, A. 2005.
ABF++: Fast and Robust Angle Based Flattening. ACM Trans.
Graph. 24, 2, 311330.

Colin de Verdire, Y. 1991. Un Principe Variationnel pour les


Empilements de Cercles. Invent. Math. 104, 655669.

Sheffer, A., Praun, E., and Rose, K. 2006. Mesh Parameterization


Methods and their Applications. Found. Trends Comput. Graph.
Vis. 2, 2, 105171.

Desbrun, M., Meyer, M., and Alliez, P. 2002. Intrinsic Parameterizations of Surface Meshes. Comp. Graph. Forum 21, 3,
209218.

Springborn, B. 2005. A Unique Representation of Polyhedral


Types. Centering via Mbius Transformations. Math. Z. 249,
513517.

Duffin, R. J. 1956. Basic Properties of Discrete Analytic Functions. Duke Math. J. 23, 335363.

Steihaug, T. 1983. The Conjugate Gradient Method and Trust


Regions in Large Scale Optimization. SIAM J. Numer. Anal. 20,
3, 626637.

Duffin, R. 1959. Distributed and Lumped Networks. J. Math.


Mech. [continued as Indiana Univ. Math. J.] 8, 793826.
Erickson, J., and Whittlesey, K. 2005. Greedy Optimal Homotopy and Homology Generators. In Proc. ACM/SIAM Symp. on
Disc. Alg., SIAM, 10381046.
Floater, M. S., and Hormann, K. 2005. Surface Parameterization: a Tutorial and Survey. In Advances in Multiresolution for
Geometric Modelling, Mathematics and Visualization. Springer,
157186.
Gu, X., and Yau, S.-T. 2003. Global Conformal Surface Parameterization. In Proc. Symp. Geom. Proc., Eurographics, 127137.
Gu, X., Gortler, S. J., and Hoppe, H. 2002. Geometry Images.
ACM Trans. Graph. 21, 3, 355361.
Jin, M., Kim, J., and Gu, X. D. 2007. Discrete Surface Ricci Flow:
Theory and Applications. In Mathematics of Surfaces 2007,
R. Martin, M. Sabin, and J. Winkler, Eds., Vol. 4647 of Lecture
Notes in Computer Science. Springer, 209232.
Klberer, F., Nieser, M., and Polthier, K. 2007. QuadCover
Surface Parameterization using Branched Coverings. Comp.
Graph. Forum 26, 3, 375384.
Kharevych, L., Springborn, B., and Schrder, P. 2006. Discrete
Conformal Mappings via Circle Patterns. ACM Trans. Graph.
25, 2, 412438.
Leibon, G. 2002. Characterizing the Delaunay Decompositions
of Compact Hyperbolic Surfaces. Geom. Topol. 6, 361391.
Lvy, B., Petitjean, S., Ray, N., and Maillot, J. 2002. Least
Squares Conformal Maps for Automatic Texture Atlas Generation. ACM Trans. Graph. 21, 3, 362371.
Lewin, L. 1981. Polylogarithms and Associated Functions. North
Holland.
Luo, F. 2004. Combinatorial Yamabe Flow on Surfaces. Commun.
Contemp. Math. 6, 765780.
Macleod, A. J. 1996. Algorithm 757: MISCFUN, a Software
Package to Compute Uncommon Special Functions. ACM Trans.
Math. Softw. 22, 3, 288301.
Mercat, C. 2001. Discrete Riemann Surfaces and the Ising Model.
Comm. in Math. Physics 218, 1, 177216.
Milnor, J. 1982. Hyperbolic Geometry: The First 150 Years. Bul.
Amer. Math. Soc. 6, 1, 924.
Pinkall, U., and Polthier, K. 1993. Computing Discrete Minimal
Surfaces and Their Conjugates. Experiment. Math. 2, 1, 1536.
Ray, N., Li, W. C., Lvy, B., Sheffer, A., and Alliez, P. 2006.
Periodic Global Parameterization. ACM Trans. Graph. 25, 4,
14601485.

Stephenson, K. 2003. Circle Packing: A Mathematical Tale.


Notices Amer. Math. Soc. 50, 11, 13761388.
Stephenson, K. 2005. Introduction to Circle Packing. Cambridge
University Press.
Tong, Y., Alliez, P., Cohen-Steiner, D., and Desbrun, M. 2007.
Designing Quadrangulations with Discrete Harmonic Forms.
In Proc. Symp. Geom. Proc., Eurographics, 201210.
Troyanov, M. 1986. Les Surfaces Euclidiennes Singularits
Coniques. Enseign. Math. (2) 32, 7994.
Yang, Y., Kim, J., Luo, F., Hu, S., and Gu, D. 2008. Optimal Surface
Parameterization Using Inverse Curvature Map. IEEE Trans.
Vis. Comp. Graph. (to appear).
Zayer, R., Lvy, B., and Seidel, H.-P. 2007. Linear Angle Based
Parameterization. In Proc. Symp. Geom. Proc., Eurographics,
135141.

Milnors Lobachevsky Function

Milnors Lobachevsky function [Milnor 1982] is defined by


Rx
L(x) = 0 log |2 sin t| dt.
Up to scale it agrees with Clausens integral Cl2 (x) =
2L(x/2) [Lewin 1981]. The function L(x) is -periodic, continuous, and odd. It is smooth except at integer multiples of
where its graph has a vertical tangent. For our purposes note that

L0 (x) = log |2 sin x|

and

L00 (x) = cot x.

(For us the absolute value signs are irrelevant because we only


consider L() for 0 .) Numerical evaluation can be performed very efficiently and with high accuracy [Macleod 1996].

Gradient of E

To show Eq. (9) for the partial derivatives of the energy E, we


show first that
ij,
jk ,
ki ) = 1
i .
jk f (
2 jk

(13)

From Eq. (8) one obtains



ij,
jk ,
ki ) = 1
jk log(2 sin
i + 12
ijk ) jk
ijk
jk f (
2 jk

j 
j
ki log(2 sin
i j log(2 sin
ki ) jk
ki + 21
kij ) jk
kij
+ 12
l jk 




l
l
j
ijk+ log 2 sin i jk
ijk+ log ki j jk
ki+ log 2 sini j k jk
kij .
= 21
jk

ki
2 sin

ij

j
(the cir ijk ) = l ki /(2 sin
ki ) = l i j /(2 sin
kij ) = R
Since l jk /(2 sin

cumcircle radius for a triangle with sides l i j , l jk , and l ki ) and

ijk +
ki +
kij ) = 0 (the angle sum is ), Eq. (13) follows.
jk (
ij,
jk ,
ki ) = +
One finds also (in view of Eq. (3)) ui f (
ij
j
ij,
jk ,
ki ) = (
k +
)/2, and hence
f (
ki

ij

ki

ij,
jk ,
ki )
ui f (

(ui
2


ijk .
+ u j + uk ) = 12

Using this, it is straightforward to derive Eq. (9) from Eq. (7).

Hessian of E

To show Eq. (10) for the Hessian of E, note that (Hess E u)i is
the derivative of the ith component of grad E in the direction u.
Since
P
bi 1
ijk ,
(grad E)i = 12
t i jk 3vi
2
b i is constant, we are done once we have shown that the
and
ijk is
first-order change in
ijk =

1
2

ki (uk ui ) + 12 cot
kij (u j ui ).
cot

(14)

and u, we start with the


To derive this relation between
j
kij / sin
ki . (Alternatively one could
sine theorem: l i j /l ki = sin
start with the cosine theorem, compare [Ben-Chen et al. 2008],
App. A.) Take the logarithm on both sides to get

and, for the first-order changes,


j

i j
ki )/2 = cot
kij
kij cot
ki
ki .
(
i j = ui + u j and
ki = uk + ui , one obtains
Since
j

kij
kij cot
ki
ki ,
(u j uk )/2 = cot
and two other such equations through cyclic permutation of
j
kij ,
ijk ,
ki , are lini jk. These three linear equations for
early dependent (they sum to zero). Two of them together with
j
ijk +
ki +
kij = 0 form a linear system of equations which

determines :

!
j
i

ijk cot
ki
cot
0
jkj 1 u j ui
A ki = 2 ui uk ,
A = cot i
.
kij
0
cot
jk
0

kij

Using the addition theorems for sine and cosine one shows
det A = 1. Applying Cramers rule then yields
!
j
1
ijk = det

ki
(u j ui ) cot

1
(ui uk )
2

kij
cot

which is Eq. (14).

l i2j = q(w i w j ) = q(w i ) 2b(w i , w j ) + q(w j ) = 2b(w i , w j )


and similarly l 2i j = 2b(w i , w j ), this is equivalent to l i2j =
ai a j l 2i j , and, using (2), to ai = 1/2 eui .

Minimal Distortion

Let M be a smooth connected oriented 2-manifold with boundary,


equipped with a Riemannian metric g. If g is a conformally
equivalent metric as in Eq. (1), then the Gaussian curvatures
e of g and g are related by Eq. (11). Thus, the metric g is flat
K, K
if u is a solution of the Poisson equation (12).
Now how can one measure the distortion caused by a conformal
change of metric? If u is constant, the new metric differs from
the old one only by a global change of scale, which we do not
consider as distortion. A reasonable measure for the distortion is
therefore the Dirichlet energy of u
R
R
D(u) = 12 M du du = 12 M g(grad u, grad u) dA

ij
ki )/2 = log sin
kij log sin
ki
(

Proof. In homogeneous coordinates w = (x, y, z), the circumcircle of 4 is described by the equation q(w) = 0 where q(w) is the
quadratic form q(w) = x 2 + y 2 + 2c xz + 2d yz + ez 2 with c, d, e
be
determined by the conditions q(w i ) = 0. Similarly, let q(w)
The circumthe quadratic form describing the circumcircle of 4.
if q(w) is up to some
circle of 4 is mapped to the cirumcircle of 4
constant factor identical to q( f (w)), i.e. q(w) = q( f (w)),
or equivalently, if the corresponding symmetric bilinear forms
w
0 ) satisfy b(w, w 0 ) = b( f (w), f (w 0 )). This is
b(w, w 0 ), b(w,
i, aj w
j ), because the w i form a
the case iff b(w i , w j ) = b(ai w
basis for R3 . But since

Circumcircle Preserving Projective Maps

in the plane with respective EuConsider two triangles 4, 4


clidean vertex coordinates pi = (x i , yi ) and pi = (
x i , yi ), and let
l i j = kpi p j k and l i j = kpi p j k be the edge lengths. Define ui
by (5) so that (2) holds. In homogeneous coordinates, pi and p i
i = (
are represented by the vectors w i = (x i , yi , 1) and w
x i , y i , 1).
come from the
The projective transformations mapping 4 to 4
linear transformations f : R3 R3 of homogeneous coordinates
i where ai R>0 .
with f (w i ) = ai w
Proposition. The projective transformation corresponding to such
a linear transformation f maps the circumcircle of 4 to the circum if and only if a1 , a2 , a3 are chosen proportional to eu1 ,
circle of 4
eu2 , eu3 .

which measures how much u changes.


Theorem. Among all conformally equivalent flat metrics g , the
ones with least distortion are obtained if u is a solution for the
Poisson equation (12) with u| M = const.
Note that this measure of distortion is symmetric: Interchanging
g and g does not change the distortion. Note also that different
choices for the constant boundary value change the solution
g only by a global scale factor, so one might as well choose
u| M = 0.
To prove the theorem, we proceed as usual in the calculus of
variations. Suppose g is flat and consider a variation of g within
the space of conformally equivalent flat metrics, i.e., a variation
of u with
u
u = 0. Then the variation of the Dirichlet energy is
R
R

) dA = M u g(grad u
, N ) ds,
D = M g(grad u, grad u
where N is the outward pointing unit normal vector field on
ranges over all smooth harmonic functions,
the boundary. As u
, N ) ranges over all smooth functions h : M R satisg(grad u
fying
R
h ds = 0.
(15)
M

Indeed, precisely for all h satisfying (15), there is a harmonic u


, N ) = h on M : This is just the Neumann boundary
with g(grad u
value problem for Laplaces equation.
So g is a critical point of the Dirichlet energy under variations
within the space of conformally equivalent flat metrics iff
R
h satisfying Eq. (15) : M u h ds = 0.
This is clearly the case if u| M is constant. To complete the
argument, suppose that u| M is not constant, so u(p1 ) > u(p2 ) for
some p1 , p2 M . Then, by continuity, there are neighborhoods
U1 , U2 of p1 and p2 such that u(q1 ) > u(q2 ) for all q1 U1 and
q2 U2 . Finally, consider a bump function h which is positive
only inside U1 , negative only inside U2 , and zero everywhere else,
and which satisfies (15) to see that u cannot be critical.

You might also like