from PIL import Image
from numpy import array
from lib import harris
im = array(Image.open('../DATA/empire.jpg').convert('L'))
harris_im = harris.compute_harris_response(im)
filtered_coords = harris.get_harris_points(harris_im, 6)
harris.plot_harris_points(im, filtered_coords)
# im2 = array(Image.open('../DATA/sf_view2.jpg').convert('L'))
# If pickle data is available, use it to display matching points instead
try:
with open(temp_folder + im_name_short + '_harris.pkl', 'rb') as f:
filtered_coords1 = pickle.load(f)
filtered_coords2 = pickle.load(f)
matches = pickle.load(f)
print('Using saved data')
# If pickle data not found, compute the interest points of each image using
# Harris corner detection; also compute the interest points that match the
# closest between the two images
except:
wid = 5
harris_im = harris.compute_harris_response(im1, 5)
filtered_coords1 = harris.get_harris_points(harris_im, wid+1)
d1 = harris.get_descriptors(im1, filtered_coords1, wid)
harris_im = harris.compute_harris_response(im2, 5)
filtered_coords2 = harris.get_harris_points(harris_im, wid+1)
d2 = harris.get_descriptors(im2, filtered_coords2, wid)
print('starting matching')
# FAIR WARNING: the following line of code takes a long time!
# This is the main reason why I have pickle data saved below (and loaded
# above) at this key point of the code.
matches = harris.match_twosided(d1, d2)
with open(temp_folder + im_name_short + '_harris.pkl', 'wb') as f:
pickle.dump(filtered_coords1, f)
