Revision history [back]

radius is the distance from the center pixel to the neighbours, like here:

with lbph, the face region gets divided into seperate patches, like this:

grid_x and grid_y are the number of patches in x and y direction (8 in this case, like the default with opencv)

for each patch, a separate (256 bin)histogram is calculated, and later concatenated to a 88256 element feature vector

[images taken from : Face Recognition with Local Binary Patterns, Timo Ahonen, Abdenour Hadid, and Matti Pietikäinen - you might want to read this..]

radius 'radius' is the distance from the center pixel to the neighbours, 'neighbours' is their count, like here:

with lbph, the face region gets divided into seperate patches, like this:

grid_x and grid_y are the number of patches in x and y direction (8 in this case, like the default with opencv)

for each patch, a separate (256 bin)histogram (2^neighbours bins)histogram is calculated, and later concatenated to a 88256

grid_x * grid_y * (2^neighbours) element feature vectorvector.

(you can already see, that increasing the neighbours count will be very expensive)

[images taken from : Face Recognition with Local Binary Patterns, Timo Ahonen, Abdenour Hadid, and Matti Pietikäinen - you might want to read this..]