Gustaf Kylberg

back to main page

LBP-based texture descriptors

The implementatuions below are all for MATLAB based on the original LBP-implementation by Marko Heikkilä and Timo Ahonen, see the following link. Both the FLBP and SLBP implementations use the same implementation of the mapping table for uniform patterns and/or rotation invariance as the original LBP implementation. This means that you need to download the function getmapping.m, also written by Marko Heikkilä and Timo Ahonen, to be able to use the flbp.m and slbp.m functions below to the full extent.

Fuzzy LBP (FLBP)

Soft LBP was introduced in 2007 [1] and in 2008 Fuzzy LBP was introduced [2] basically being the same LBP-based method where a pixel position can contribute to several bins in the histogram of occurring binary codes. Contribution (or belongingness to the different codes) sums up to 1 for each pixel position.

A naive implementation of FLBP would be to compute belongingness to each bin for each pixel position/local neighborhood. However, if a local neighborhood has high enough contrast the number of possible bins that will receive a contribution is restricted. By also restricting the belongings computations to only the relevant bins a faster implementation can be written. This is the basis of the FLBP implementation below.

Shift LBP (SLBP)

Shift LBP is an attempt to capture some of the fuzzy nature of the local neighborhoods without the need for computing belongingness to each bin.


In addition getmapping.m by Marko Heikkilä and Timo Ahonen is also necessary for full functionality of flbp.m and slbp.m.


[1] Ahonen, T. & Pietikäinen, M. Soft histograms for local binary patterns In Proceedings of the Finnish signal processing symposium, (FINSIG 2007),, 2007, 1-4.

[2] Iakovidis, D. K.; Keramidas, E. G. & Maroulis, D. Fuzzy Local Binary Patterns for Ultrasound Texture Characterization Image Analysis and Recognition, Springer Berlin / Heidelberg,, 2008, 5112, 750-759 [doi]


Shift LBP (SLBP)

  1. Florindo, J. B. & Bruno, O. M.
    Fractal descriptors based on the probability dimension: A texture analysis and classification approach
    Pattern Recognition Letters, 2014, 42, 107 - 114.
  2. Kylberg G., and Sintorn I.-M.
    Evaluation of Noise Robustness for Local Binary Pattern Descriptors in Texture Classification
    EURASIP Journal on Image and Video Processing, 2013, 17.
    [abstract] [doi] [pdf]