selected block normalization method is invalid

[Landay7]

When i try to run python detector.py I have next problem

Can you help, please?

[Windaway]

Remove normalize. Skimage will remove the features.

[zhixinma]

The definition of function hog is as follows

def hog(image, orientations=9, pixels_per_cell=(8, 8), cells_per_block=(3, 3),
        block_norm='L1', visualise=False, transform_sqrt=False,
        feature_vector=True, normalise=None)

so removing normalize and visualise will works, specifying the parameter name will works too

[Weipeilang]

When I was running the code, I found that it didn't work very well. How can I make it work better

[qlerebours]

@Weipeilang i had the same issue but I think I found the answer:

When they work on the human detector, they worked with scikit-image in v0.12.x: Doc in v0.12
In the linked version, the hog method didn't have the current block_norm param, included in the v0.13.x
Current doc version 0.15.x

This is why we get this error message. To fix this, you can simply replace hog(im, orientations, pixels_per_cell, cells_per_block, visualize, normalize) by hog(im, orientations, pixels_per_cell, cells_per_block, visualize=visualize, transform_sqrt=normalize) to prevent visualize value to be assigned to block_norm value (causing the error)

@BUPTLdy could you maybe update the README to specify the version of lib used ?

[shahsamkit73]

I am getting this error during execution....