def test_stitching_color():
im = color.gray2rgb(data.moon())
l = 256
n_rows = 1
n_cols = im.shape[1] // l
init_i, init_j = 0, 0
imgs = np.empty((n_rows, n_cols, l, l, 3))
overlap_h = [5, 50]
i = 0
for j in range(n_cols):
sub_im = im[init_i:init_i + l, init_j:init_j + l]
imgs[i, j] = sub_im
init_j += l - overlap_h[1]
stitch = register_tiles(imgs,
n_rows,
n_cols,
overlap_global=0.2,
pad=l // 2)
delta = im[:l, :stitch.shape[1]].astype(np.float) - stitch.astype(np.float)
assert np.all(delta == 0)
def test_stitching_two_rows():
im = data.moon()
l = 256
# two rows
n_rows = 2
n_cols = im.shape[1] // l
init_i, init_j = 0, 0
overlap_h = [5, 50]
overlap_v = 40
imgs = np.empty((n_rows, n_cols, l, l))
for i in range(n_rows):
for j in range(n_cols):
sub_im = im[init_i:init_i + l, init_j:init_j + l]
imgs[i, j] = sub_im
init_j += l - overlap_h[1]
init_j = 0
init_i += l - overlap_v
stitch = register_tiles(imgs,
n_rows,
n_cols,
overlap_global=0.2,
pad=l // 2)
delta = im[:stitch.shape[0], :stitch.shape[1]].astype(
np.float) - stitch.astype(np.float)
print(delta.mean())
assert np.all(delta == 0)
def test_stitching_one_row():
im = data.moon()
l = 256
n_rows = 1
n_cols = im.shape[1] // l
init_i, init_j = 0, 0
imgs = np.empty((n_rows, n_cols, l, l))
overlap_h = [5, 50]
i = 0
for j in range(n_cols):
sub_im = im[init_i:init_i + l, init_j:init_j + l]
imgs[i, j] = sub_im
init_j += l - overlap_h[1]
stitch = register_tiles(imgs,
n_rows,
n_cols,
overlap_global=0.2,
pad=l // 2)
delta = im[:l, :stitch.shape[1]].astype(np.float) - stitch.astype(np.float)
assert np.all(delta == 0)
stitch = register_tiles(imgs,
n_rows,
n_cols,
overlap_global=0.2,
pad=l // 2,
blending=False)
delta = im[:l, :stitch.shape[1]].astype(np.float) - stitch.astype(np.float)
assert np.all(delta == 0)
# local_overlap
stitch = register_tiles(imgs,
n_rows,
n_cols,
overlap_global=0.5,
pad=l // 2,
overlap_local={(0, 1): [0, 45]})
delta = im[:l, :stitch.shape[1]].astype(np.float) - stitch.astype(np.float)
assert np.all(delta == 0)
def modify_content(n_cl, n_rows, n_cols, overlap, estimate, image_string):
tiles = untile_images(image_string, n_rows, n_cols)
if estimate is not None and len(estimate) > 0:
overlap = []
for line in estimate:
overlap.append(1.1 * line['length'] / tiles.shape[3])
canvas = register_tiles(tiles,
n_rows,
n_cols,
overlaps=overlap,
pad=100)
return array_to_data_url(canvas)
def modify_content(n_cl, n_rows, n_cols, overlap, estimate, image_string):
if image_string is None:
raise PreventUpdate
tiles = untile_images(image_string, n_rows, n_cols)
if estimate is not None and len(estimate) > 0:
overlap_dict = _sort_props_lines(estimate, tiles.shape[2],
tiles.shape[3], n_cols)
else:
overlap_dict = None
canvas = register_tiles(tiles, n_rows, n_cols,
overlap_global=overlap,
overlap_local=overlap_dict,
pad=np.max(tiles.shape[2:])//2,
blending=True)
return array_to_data_url(canvas)