michael/stable-diffusion-webui
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

fix entropy point calculation

Browse source
This commit is contained in:
captin411 2022-10-19 13:44:59 -07:00 committed by GitHub
parent 087609ee18
commit 41e3877be2
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 19 additions and 15 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

34
modules/textual_inversion/preprocess.py
View file

@ -196,9 +196,9 @@ def image_focal_points(im):
points = cv2.goodFeaturesToTrack( points = cv2.goodFeaturesToTrack(
np_im, np_im,
maxCorners=50, maxCorners=100,
qualityLevel=0.04, qualityLevel=0.04,
minDistance=min(grayscale.width, grayscale.height)*0.05, minDistance=min(grayscale.width, grayscale.height)*0.07,
useHarrisDetector=False, useHarrisDetector=False,
) )
@ -218,28 +218,32 @@ def image_focal_points(im):
def image_entropy_point(im, crop_width, crop_height): def image_entropy_point(im, crop_width, crop_height):
img = im.copy() landscape = im.height < im.width
# just make it easier to slide the test crop with images oriented the same way portrait = im.height > im.width
if (img.size[0] < img.size[1]): if landscape:
portrait = True move_idx = [0, 2]
img = img.rotate(90, expand=1) move_max = im.size[0]
elif portrait:
move_idx = [1, 3]
move_max = im.size[1]
e_max = 0 e_max = 0
crop_current = [0, 0, crop_width, crop_height] crop_current = [0, 0, crop_width, crop_height]
crop_best = crop_current crop_best = crop_current
while crop_current[2] < img.size[0]: while crop_current[move_idx[1]] < move_max:
crop = img.crop(tuple(crop_current)) crop = im.crop(tuple(crop_current))
e = image_entropy(crop) e = image_entropy(crop)
if (e_max < e): if (e > e_max):
e_max = e e_max = e
crop_best = list(crop_current) crop_best = list(crop_current)
crop_current[0] += 4 crop_current[move_idx[0]] += 4
crop_current[2] += 4 crop_current[move_idx[1]] += 4
x_mid = int(crop_best[0] + crop_width/2)
y_mid = int(crop_best[1] + crop_height/2)
x_mid = int((crop_best[2] - crop_best[0])/2)
y_mid = int((crop_best[3] - crop_best[1])/2)
return { return {
'x': x_mid, 'x': x_mid,
@ -250,7 +254,7 @@ def image_entropy_point(im, crop_width, crop_height):
def image_entropy(im): def image_entropy(im):
# greyscale image entropy # greyscale image entropy
band = np.asarray(im.convert("L")) band = np.asarray(im.convert("1"))
hist, _ = np.histogram(band, bins=range(0, 256)) hist, _ = np.histogram(band, bins=range(0, 256))
hist = hist[hist > 0] hist = hist[hist > 0]
return -np.log2(hist / hist.sum()).sum() return -np.log2(hist / hist.sum()).sum()