LBP FEATURE EXTRACTION METHOD

I have extracted LBP features of the image dataset.Can anybody xplain me do I have done in the right way or not.

Description: I have 2 folders one is benign( 428) and other one is malignant(227). So I have extract LBP features of both the folders. Total no. of images is 655 (i.e. 428 + 227).

I have used

F = extractLBPFeatures(I);

in MATLAB2016b

Can anybody explain what is 655 X 59 MEANS OVER HERE. 655 is the total no. of images , what is 59 ?

Answer 1

[Bhavik Patel]

It is my understanding that you want to know about the number of columns in the output matrix returned by extractLBPFeatures function.

The extractLBPFeatures function computes uniform local binary patterns for grayscale image. By default, the function uses 8 neighbours (P = 8), and it does not encode rotation information, i.e., Upright argument is set to true.

By default, the number of output bins are (P×P-1)+3. For P = 8, this value is (8×7)+3 = 59. So, the number 59 represents the number of bins in the LBP histogram.

To know more, see documentation of extractLBPFeatures, and refer to the “References” section.

Hope this helps!