Local binary patterns: Difference between revisions

Revision as of 02:56, 20 June 2010

Revision as of 18:59, 31 July 2010

Line 1:

'''Local Binary Patterns''' (abbreviated LBP) is a type of feature used for classification in [[computer vision]]. ItLBP was first described by Ojala, Pietikäinen and Harwood in 1996 .<ref>T. Ojala, M. Pietikäinen, and D. Harwood (1996), "A Comparative Study of Texture Measures with Classification Based on Feature Distributions", Pattern Recognition, vol. 29, pp. 51-59. This was the first article to describe LBP.</ref> , andIt has since been found to be a powerful feature for texture classification.; Itit has alsofurther been founddetermined that when LBP is combined with the [[Histogram of oriented gradients]] (HOG) classifier, it yields the best classifier of humans (i.e. person vs. non-person), among the classifiers usually considered in academic literature<ref>"An HOG-LBP Human Detector with Partial Occlusion Handling", Xiaoyu Wang, Tony X. Han, Shuicheng Yan, ICCV 2009</ref>.

{{Wikify|date=April 2010}}

==Concept==

Local Binary Patterns (abbreviated LBP) is a type of feature used for classification in computer vision. It was first described by Ojala, Pietikäinen and Harwood in 1996 <ref>T. Ojala, M. Pietikäinen, and D. Harwood (1996), "A Comparative Study of Texture Measures with Classification Based on Feature Distributions", Pattern Recognition, vol. 29, pp. 51-59</ref> , and has since been found to be a powerful feature for texture classification. It has also been found that when combined with the HOG classifier, it yields the best classifier of humans (i.e. person vs. non-person), among the classifiers usually considered in academic literature<ref>"An HOG-LBP Human Detector with Partial Occlusion Handling", Xiaoyu Wang, Tony X. Han, Shuicheng Yan, ICCV 2009</ref>.

The LBP feature vector, in its simplest form, is created in the following manner:

==Basic Idea==

LBP feature vector, in its simplest form, is created in the following manner:

* Divide the examined window to cells (e.g. 16x16 pixels for each cell).

Line 13 ⟶ 11:

* Concatenate normalized histograms of all cells. This gives the feature vector for the window.

The feature vector now can be processed using the [[Support vector machine]] or some other machine-learning algorithm, to produce a classifier.

==References==