def create_local_homography_object(bandwidth, magnitude, lambda_):
"""
Helper function for creating WeightedLocalHomography objects
"""
H = WeightedLocalHomography(SqExpWeightingFunction(bandwidth, magnitude))
H.regularization_lambda = lambda_
return H
def get_homography_estimate(filename):
#
# Conventions:
# a_i, b_i
# are variables in image space, units are pixels
# a_w, b_w
# are variables in world space, units are meters
#
print '\n========================================'
print ' File: ' + filename
print '========================================\n'
from skimage.io import imread
from skimage.color import rgb2gray
im = imread(filename)
im = rgb2gray(im)
detections = get_tag_detections(im, cache_tag=filename)
print ' %d tags detected.' % len(detections)
#
# Sort detections by distance to center
#
c_i = np.array([im.shape[1], im.shape[0]]) / 2.
dist = lambda p_i: np.linalg.norm(p_i - c_i)
closer_to_center = lambda d1, d2: int(dist(d1.c) - dist(d2.c))
detections.sort(cmp=closer_to_center)
from tag36h11_mosaic import TagMosaic
tag_mosaic = TagMosaic(0.0254)
mosaic_pos = lambda det: tag_mosaic.get_position_meters(det.id)
det_i = np.array([ d.c for d in detections ])
det_w = np.array([ mosaic_pos(d) for d in detections ])
#
# Improvisation on the classic calibration procedure. We
# obtain our initial homography estimate from 9 detections
# at the center. This is better because the distortion is
# minimal at the center.
#
det_i9 = det_i[:9]
det_w9 = det_w[:9]
from projective_math import WeightedLocalHomography, UnitWeightingFunction
H_estimator = WeightedLocalHomography(UnitWeightingFunction())
for s, t in zip(det_w9, det_i9):
H_estimator.add_correspondence(s, t)
from tupletypes import Correspondence
H = H_estimator.get_homography_at(det_w9[0])
corrs = [ Correspondence(w, i) for w, i in zip(det_w, det_i) ]
return HomographyInfo(corrs, H, im.shape)
