Seam-aware Decimater

This project implements the seam-aware decimation portion of our SIGGRAPH Asia 2017 paper Seamless: Seam erasure and seam-aware decoupling of shape from mesh resolution. Seam-aware decimater simplifies a mesh while preserving its UV's boundary. It allows the same texture to be used across all decimation levels—notably along seams.

Our seam-aware decimation allows seamless texture reuse at all decimation levels (here, approximately 1%). Seams on the original model are shown in purple. Parameterizations are shown inset. Garland and Heckbert [1998] (implemented by MeshLab [Cignoni et al. 2008]) do not preserve seams precisely, leading to artifacts in the texture. Red areas near seams in the inset parameterization indicate this deviation in the parametric domain. Maya [2017] prevents decimation of seams entirely, leading to suboptimal allocation of mesh vertices. <img src = "results/blub_representative.png" width="100%">

Requirements

This project uses C++ 11, and it depends on:

• libigl (git clone https://github.com/libigl/libigl.git --recursive)
• eigen (e.g. brew install eigen)

Compile this project

mkdir build
cd build
cmake -DCMAKE_BUILD_TYPE=Release ..
make

Run this project

./decimater ../models/animal.obj percent-vertices 50
./decimater ../models/animal.obj num-vertices 1000

Note: this library only works on triangle mesh.

Various strictness

Seamless criteria is even stricter than UV shape preservation. I add a "strictness" parameter to satisfy different requirements. Strictness could be 0, 1, or 2. Strictness 0 means no UV shape is preserved, but UV parameters is still part of the metrics. Strictness 1 means UV shape is preserved. Strictness 2 means seam aware decimation which also considers the length ratio criteria. It works like this:

./decimater ../models/animal.obj percent-vertices 50 --strict 1
./decimater ../models/animal.obj num-vertices 1000 --strict 1

The default strictness is 2.

Example

The Animal model is decimated to 3% of its original number of vertices. The boundary of its UV parameterization stays. <img src = "results/extreme_decimation.001.png" width="100%"> <img src = "results/strictness.png" width="100%">