How to extract a feature vector of size 256 from Local Binary pattern (LBP ) Image?

I had done enough literature on LBP but i guess i still fail to understand it completely. According to my understanding when LBPH is applied on an image it returns a feature vector of size 1x256 i.e. the feature vector contians the frequency of gray levels (from 0-255) in the image. Ive tried many LBP codes i found online and i dont understand what it returns.

Simple what i want is a feature vector of size 1x256 when LBP is applied on an image.

Please guide me and provide the source code (code is much needed)

ive followed this code by ImageAnalyst : https://in.mathworks.com/matlabcentral/answers/15774-local-binary-pattern-lbp#comment_687497 but i dont know what is the final feature vector

Answer 1

[Gayathri]

Hello @Unqua Laraib,

LBP is a simple and grayscale invariant texture descriptor measure for classification. The length of the extracted feature depends on the size of the image and does not always have to be a 256-length vector.

MATLAB provides a function “extractLBPFeatures” to find the LBP features without coding it explicitly. We can obtain the LBP features a follows.

features = extractLBPFeatures(I)

where I is the input image.

For more information on the “extractLBPFeatures” function and examples associated with it, please refer to the following link.https://www.mathworks.com/help/vision/ref/extractlbpfeatures.html

Hope you find this information helpful.