3
$\begingroup$

I'm currently looking into Mesh Simplification to batch-simplify a large amount of .obj files. It is necessary for these algorithms to preserve the visual representation of a mesh as well as possible.

During research I've stumbled over Garland & Heckberts Surface Simplification Using Quadric Error Metrics. Most of the frameworks I've inspected are making use of this algorithm for Mesh Simplification.

My question is, why are most frameworks not making use of seemingly more advanced algorithms e.g. Simplifying Surfaces with Color and Texture using Quadric Error Metrics or comparable? Is the Quadric Error Metric currently state of the art concerning simplification algorithms?

$\endgroup$
2
$\begingroup$

According to Polygon-Mesh-Processing, section 7.2.2. paragraph "Error quadrics"

Error quadrics have major advantages regarding memory consumption and computational efficiency: each vertex $x$ stores a symmetric 4x4 matrix $Q$ only, and the error can efficiently be computed in constant time as $x^T Q x$ - no matter how many planes are associated to vertex $x$ during decimation. Because of this, error quadrics are one of the most frequently employed techniques in mesh decimation.

The author of the reference I mentioned specifically refers to the paper you mention in your question. My interpretation is that more recent methods might be more sophisticated but resource-wise quadric errors are still better.

Also I'd add is relatively easy to implement anyway.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.