Progressive Meshes
ABSTRACT
Highly detailed geometric models are rapidly becoming common-
place in computer graphics. These models, often represented as
complex triangle meshes, challenge rendering performance, trans-
mission bandwidth, and storage capacities. This paper introduces
the progressive mesh (PM) representation, a new scheme for storing
and transmitting arbitrary triangle meshes. This efficient, loss-
less, continuous-resolution representation addresses several practi-
cal problems in graphics: smooth geomorphing of level-of-detail
approximations, progressive transmission, mesh compression, and
selective refinement.
In addition, we present a new mesh simplification procedure for
constructing a PM representation from an arbitrary mesh. The goal
of this optimization procedure is to preserve not just the geometry
of the original mesh, but more importantly its overall appearance
as defined by its discrete and scalar appearance attributes such as
material identifiers, color values, normals, and texture coordinates.
We demonstrate construction of the PM representation and its ap-
plications using several practical models.
CR Categories and Subject Descriptors: I.3.5 [Computer Graphics]:
Computational Geometry and Object Modeling - surfaces and object repre-
sentations.
Additional Keywords: mesh simplification, level of detail, shape interpo-
lation, progressive transmission, geometry compression.
1 INTRODUCTION
Highly detailed geometric models are necessary to satisfy a grow-
ing expectation for realism in computer graphics. Within traditional
modeling systems, detailed models are created by applying ver-
satile modeling operations (such as extrusion, constructive solid
geometry, and freeform. deformations) to a vast array of geometric
primitives. For efficient display, these models must usually be tes-
sellated into polygonal approximations—meshes. Detailed meshes
are also obtained by scanning physical objects using range scanning
systems [5]. In either case, the resulting complex meshes are ex-
pensive to store, transmit, and render, thus motivating a number of
practical problems:
Email:
Web: http://www.research.microsoft.com/research/graphics/hoppe/
Mesh simplification: The meshes created by modeling and scan-
ning systems are seldom optimized for rendering efficiency, and
can frequently be replaced by nearly indistinguishable approx-
imations with far fewer faces. At present, this process often
requires significant user intervention. Mesh simplification tools
can hope to automate this painstaking task, and permit the port-
ing of a single model to platforms of varying performance.
Level-of-detail (LOD) approximation: To further improve ren-
dering performance, it is common to define several versions of a
model at various levels of detail [3, 8]. A detailed mesh is used
when the object is close to the viewer, and coarser approxima-
tions are substituted as the object recedes. Since instantaneous
switching between LOD meshes may lead to perceptible “pop-
ping”, one would like to construct smooth visual transitions,
geomorphs, between meshes at different resolutions.
Progressive transmission: When a mesh is transmitted over a
communication line, one would like to show progressively better
approximations to the model as data is incrementally received.
One approach is to transmit successive LOD approximations,
but this requires additional transmission time.
Mesh compression: The problem of minimizing the storage
space for a model can be addressed in two orthogonal ways.
One is to use mesh simplification to reduce the number of faces.
The other is mesh compression: minimizing the space taken to
store a particular mesh.
Selective refinement: Each mesh in a LOD representation cap-
tures the model at a uniform. (view-independent) level of detail.
Sometimes it is desirable to adapt the level of refinement in se-
lected regions. For instance, as a user flies over a terrain, the
terrain mesh need be fully detailed only near the viewer, and
only within the field of view.
In addressing these problems, this paper makes two major con-
tributions. First, it introduces the progressive mesh (PM) repre-
sentation. In PM form, an arbitrary mesh