Intensity-curvature functional-based digital high-pass filters

摘要

The intensity-curvature functional (ICF) of a model polynomial function is defined on a pixel-by-pixel basis by the ratio between the intensity-curvature term before interpolation and the intensity-curvature term after interpolation. Through the comparison with the traditional high-pass filter (HPF), this work presents evidence that the ICFs of three model polynomial functions can be tuned as HPFs. The evidence consists of the mathematical characterization of the ICF-based HPFs, qualitative comparisons in magnetic resonance imaging (MRI) of the human brain, and the determination of the finite impulse response (FIR) of the filters. The ICF-based HPFs can remove periodic noise in the low-frequency band.