Mercurial > hg > nsaunier > traffic-intelligence

comparison python/cvutils.py @ 893:ff92801e5c54

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
updated hog to scikit-image 0.13 (needed to add a block_norm attribute in classifier.cfg)
author Nicolas Saunier <nicolas.saunier@polymtl.ca>
date Tue, 30 May 2017 16:10:18 -0400
parents 1fdafa9f6bf4
children 0c1fed9e8862
comparison
equal deleted inserted replaced
892:f766fe0995f4 893:ff92801e5c54
603 603
604 if skimageAvailable: 604 if skimageAvailable:
605 from skimage.feature import hog 605 from skimage.feature import hog
606 from skimage import color, transform 606 from skimage import color, transform
607 607
608 def HOG(image, rescaleSize = (64, 64), orientations=9, pixelsPerCell=(8, 8), cellsPerBlock=(2, 2), visualize=False, normalize=False): 608 def HOG(image, rescaleSize = (64, 64), orientations=9, pixelsPerCell=(8,8), cellsPerBlock=(2,2), blockNorm='L1', visualize=False, normalize=False):
609 bwImg = color.rgb2gray(image) 609 bwImg = color.rgb2gray(image)
610 inputImg = transform.resize(bwImg, rescaleSize) 610 inputImg = transform.resize(bwImg, rescaleSize)
611 features = hog(inputImg, orientations, pixelsPerCell, cellsPerBlock, visualize, normalize) 611 features = hog(inputImg, orientations, pixelsPerCell, cellsPerBlock, blockNorm, visualize, normalize)
612 if visualize: 612 if visualize:
613 from matplotlib.pyplot import imshow, figure, subplot 613 from matplotlib.pyplot import imshow, figure, subplot
614 hogViz = features[1] 614 hogViz = features[1]
615 features = features[0] 615 features = features[0]
616 figure() 616 figure()
618 imshow(inputImg) 618 imshow(inputImg)
619 subplot(1,2,2) 619 subplot(1,2,2)
620 imshow(hogViz) 620 imshow(hogViz)
621 return float32(features) 621 return float32(features)
622 622
623 def createHOGTrainingSet(imageDirectory, classLabel, rescaleSize = (64, 64), orientations=9, pixelsPerCell=(8, 8), cellsPerBlock=(2, 2), visualize=False, normalize=False): 623 def createHOGTrainingSet(imageDirectory, classLabel, rescaleSize = (64,64), orientations=9, pixelsPerCell=(8,8), blockNorm='L1', cellsPerBlock=(2, 2), visualize=False, normalize=False):
624 inputData = [] 624 inputData = []
625 for filename in listdir(imageDirectory): 625 for filename in listdir(imageDirectory):
626 img = imread(imageDirectory+filename) 626 img = imread(imageDirectory+filename)
627 features = HOG(img, rescaleSize, orientations, pixelsPerCell, cellsPerBlock, visualize, normalize) 627 features = HOG(img, rescaleSize, orientations, pixelsPerCell, cellsPerBlock, blockNorm, visualize, normalize)
628 inputData.append(features) 628 inputData.append(features)
629 629
630 nImages = len(inputData) 630 nImages = len(inputData)
631 return array(inputData, dtype = float32), array([classLabel]*nImages) 631 return array(inputData, dtype = float32), array([classLabel]*nImages)
632 632