feature_extraction.py

import numpy
import cv2
from segmentation import region_from_segment
from opencv_utils import background_color

FEATURE_DATATYPE=   numpy.float32
#FEATURE_SIZE is defined on the specific feature extractor instance
FEATURE_DIRECTION=  1 #horizontal - a COLUMN feature vector
FEATURES_DIRECTION= 0 # vertical - ROWS of feature vectors

class FeatureExtractor( object ):
    '''given a list of segments, returns a list of feature vectors'''
    def extract( self, image, segments ):
        raise NotImplementedError()

class SimpleFeatureExtractor( FeatureExtractor ):
    def __init__(self, feature_size=10, stretch=False):
        self.feature_size= feature_size
        self.stretch=stretch

    def extract(self, image, segments):
        image= cv2.cvtColor(image,cv2.COLOR_BGR2GRAY)
        fs= self.feature_size
        bg= background_color( image )
 
        regions=numpy.ndarray( shape=(0,fs), dtype=FEATURE_DATATYPE )
        for segment in segments:
            region= region_from_segment( image, segment )
            if self.stretch:
                region = cv2.resize(region, (fs,fs) )
            else:
                x,y,w,h= segment
                proportion= float(min(h,w))/max(w,h)
                new_size= (fs, int(fs*proportion)) if min(w,h)==h else (int(fs*proportion), fs)
                region = cv2.resize(region, new_size)
                s= region.shape
                newregion= numpy.ndarray( (fs,fs), dtype= region.dtype )
                newregion[:,:]= bg
                newregion[:s[0],:s[1]]= region
                region=newregion
            regions= numpy.append( regions, region, axis=0 )
        regions.shape=( len(segments), fs**2 )
        return regions