from register.metrics import metric
from register.samplers import sampler
from register import register
from register.visualize import plot
# Form some test data (lena, lena rotated 20 degrees)
image = misc.lena()
template = nd.rotate(image, 20, reshape=False)
# Form the affine registration instance.
affine = register.Register(
model.Affine,
metric.Residual,
sampler.CubicConvolution
)
# Coerce the image data into RegisterData.
image = register.RegisterData(image).downsample(2)
template = register.RegisterData(template).downsample(2)
# Register.
step, search = affine.register(
image,
template,
verbose=True,
)
# Call the debug tool "searchInspector"
plot.searchInspector(search)
metric.Residual,
sampler.CubicConvolution
)
fullSearch = []
# Image pyramid registration can be executed like so:
pHat = None
for factor in [30., 20. , 10., 5., 2., 1.]:
if pHat is not None:
scale = downImage.coords.spacing / factor
# FIXME: Find a nicer way to do this.
pHat = model.Affine.scale(pHat, scale)
downImage = image.downsample(factor)
downTemplate = template.downsample(factor)
step, search = affine.register(
downImage,
downTemplate,
p=pHat,
verbose=True
)
pHat = step.p
fullSearch.extend(search)
plot.searchInspector(fullSearch)
