feat: extract attribute canny_high_threshold

This commit is contained in:
Manuel Schmid
2024-01-22 19:08:54 +01:00
parent ec486443ea
commit 2d8ca41ce5
4 changed files with 16 additions and 15 deletions
+8 -8
View File
@@ -3,25 +3,25 @@ import numpy as np
import modules.advanced_parameters as advanced_parameters
def centered_canny(x: np.ndarray, canny_low_threshold):
def centered_canny(x: np.ndarray, canny_low_threshold, canny_high_threshold):
assert isinstance(x, np.ndarray)
assert x.ndim == 2 and x.dtype == np.uint8
y = cv2.Canny(x, int(canny_low_threshold), int(advanced_parameters.canny_high_threshold))
y = cv2.Canny(x, int(canny_low_threshold), int(canny_high_threshold))
y = y.astype(np.float32) / 255.0
return y
def centered_canny_color(x: np.ndarray, canny_low_threshold):
def centered_canny_color(x: np.ndarray, canny_low_threshold, canny_high_threshold):
assert isinstance(x, np.ndarray)
assert x.ndim == 3 and x.shape[2] == 3
result = [centered_canny(x[..., i], canny_low_threshold) for i in range(3)]
result = [centered_canny(x[..., i], canny_low_threshold, canny_high_threshold) for i in range(3)]
result = np.stack(result, axis=2)
return result
def pyramid_canny_color(x: np.ndarray, canny_low_threshold):
def pyramid_canny_color(x: np.ndarray, canny_low_threshold, canny_high_threshold):
assert isinstance(x, np.ndarray)
assert x.ndim == 3 and x.shape[2] == 3
@@ -31,7 +31,7 @@ def pyramid_canny_color(x: np.ndarray, canny_low_threshold):
for k in [0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0]:
Hs, Ws = int(H * k), int(W * k)
small = cv2.resize(x, (Ws, Hs), interpolation=cv2.INTER_AREA)
edge = centered_canny_color(small, canny_low_threshold)
edge = centered_canny_color(small, canny_low_threshold, canny_high_threshold)
if acc_edge is None:
acc_edge = edge
else:
@@ -54,11 +54,11 @@ def norm255(x, low=4, high=96):
return x * 255.0
def canny_pyramid(x, canny_low_threshold):
def canny_pyramid(x, canny_low_threshold, canny_high_threshold):
# For some reasons, SAI's Control-lora Canny seems to be trained on canny maps with non-standard resolutions.
# Then we use pyramid to use all resolutions to avoid missing any structure in specific resolutions.
color_canny = pyramid_canny_color(x, canny_low_threshold)
color_canny = pyramid_canny_color(x, canny_low_threshold, canny_high_threshold)
result = np.sum(color_canny, axis=2)
return norm255(result, low=1, high=99).clip(0, 255).astype(np.uint8)