teaser

Bilinear Accelerated Filter Approximation

We approximate filters like in Cardinality-Constrained Texture Filtering (CCTF). Whereas CCTF places no restrictions on which texels to read, we add an additional structural constraint that the texels can be read using a bilinear sampler. This provides better cache coherency and hardware acceleration of the texture filtering, which improves speed, with only a small decrease in accuracy.

download pdf download preprint

BibTeX

@article{Manson:2014:BAFA,
  author  = {Josiah Manson and
             Scott Schaefer},
  title   = {Bilinear Accelerated Filter Approximation},
  journal = {Computer Graphics Forum (Proceedings of EGSR)},
  year    = {2014},
  volume  = {33},
  number  = {4},
  pages   = {33--40},
}

Abstract

Our method approximates exact texture filtering for arbitrary scales and translations of an image while taking into account the performance characteristics of modern GPUs. Our algorithm is fast because it accesses textures with a high degree of spatial locality. Using bilinear samples guarantees that the texels we read are in a regular pattern and that we use a hardware accelerated path. We control the texel weights by manipulating the u, v parameters of each sample and the blend factor between the samples. Our method is similar in quality to Cardinality-Constrained Texture Filtering but runs two times faster.

Supplemental

download ppt egsr2014.ppt

The presentation that I gave at EGSR 2014

download zip resample_bilinear.zip

OpenGL resampling code using GLSL shaders.

download zip resample_bi_sw.zip

Resampling code using plain old C++ on the CPU.

download zip bilinear_optimize.zip

Code used to perform actual optimization of sample placement and coefficients.

download zip results.zip

Tables of the results of my optimizations for different table sizes and methods.

download zip images.zip

Comparison images resulting from this algorithm, CCTF, and trilinear filtering.

License

The images, executables, and code supplied are from the web page http://josiahmanson.com. These materials are free to use for non-commercial purposes. Any works that use materials from this web page should acknowledge Josiah Manson and the paper Bilinear Accelerated Filter Approximation. For commercial use, please contact Josiah Manson (josiahmanson@gmail.com).