capmoe.cv

capmoe.cv.bof

synopsis:Create BoF representation from features & visual words

class capmoe.cv.bof.BoFMaker(visualwords, index_filepath=None, algorithm=u'kdtree', loglevel=u'WARNING')

Bases: object

Create BoF representation from features & visual words.

Since ANN algorithm is used internally, the result BoF representation is approximation.

Index creation is wrapped by this class.

__init__(visualwords, index_filepath=None, algorithm=u'kdtree', loglevel=u'WARNING')

Create or load index of visual words.

Parameters:

• visualwords (M x len(feature) numpy.ndarray, where each element is numpy.float32) – 2D array of base vectors
• index_filepath – If not None, index is not created but load from specified file. The file must be created by BoFMaker.save()
• algorithm – passed to pyflann.FLANN().build_index

make(features, norm_order=None)

Create BoF representation of features.

Parameters:

• features (N x len(feature) numpy.ndarray, where each element is numpy.float32) – 2D array of feature vectors
• norm_order – 1 for L1-norm, 2 for L2-norm, ... Histogram is not normalized when this is None.

static meta_filepath(index_filepath)

Generate path to meta data file

save(filepath)

Save index of visual words to filepath

capmoe.cv.capdetector

synopsis:Provides function to detect a beer cap from an image

capmoe.cv.capdetector.capdetector(imgpath, max_candidates, loglevel=u'WARNING')

Detect circles from an image

Return type:[{‘x’: <cap circle center x>, ‘y’: <cap circle center y>, ‘r’: <cap circle radius>}, ...]

capmoe.cv.visualwords

synopsis:Creates visual words (codebook) from features of images

capmoe.cv.visualwords.visualwords_union(features, loglevel=u'WARNING')

Create simple visual words just by making union of feature vectors

Not suitable for large number of features.

Parameters:features –

feature vectors. Iterable of 2-D numpy.ndarray. Example:

Return type:2-D numpy.ndarray