Spaces:
Running
on
CPU Upgrade
Running
on
CPU Upgrade
File size: 3,347 Bytes
57a1960 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 |
import cv2
import numpy as np
from PIL import Image
# marker choices
COLORS = [(246, 195, 203), (112, 221, 208)]
MARKERS = [1, 5]
def blend_seg(input_image, segmented_image, dot_locations, dot_labels, contour_color=(63, 126, 174)):
input_image = np.array(input_image)
segmented_image = np.array(segmented_image)
# Create a mask for the foreground (non-transparent) pixels
foreground_mask = segmented_image[:, :, 3] > 0
# Create a mask for the background (transparent) pixels
background_mask = ~foreground_mask
# Darken the background pixels
darkened_background = input_image.copy()
darkened_background[background_mask] = darkened_background[background_mask] * 0.52 # Adjust the multiplier as needed to control darkness
# Create an empty mask for the boundary
boundary_mask = np.zeros_like(segmented_image[:, :, 3], dtype=np.uint8)
solid_boundary_mask = np.zeros_like(segmented_image[:, :, 3], dtype=np.uint8)
# Find the contour of the segmented region
contours, _ = cv2.findContours(
segmented_image[:, :, 3], cv2.RETR_LIST, cv2.CHAIN_APPROX_SIMPLE
)
# Draw the boundary on the boundary mask
cv2.drawContours(boundary_mask, contours, -1, (255), thickness=6)
cv2.drawContours(solid_boundary_mask, contours, -1, (255), thickness=2)
blur_mask = cv2.GaussianBlur(boundary_mask, (0, 0), sigmaX=4)
# Create a mask for the contour region
contour_region_mask = (blur_mask > 0)
# Blend the contour color with the existing pixel colors
result_image = darkened_background.copy()
mask_weight = 0.9 * blur_mask[contour_region_mask, None]/255
result_image[contour_region_mask] = (
darkened_background[contour_region_mask] * (1-mask_weight) + np.array(contour_color) * mask_weight
).astype(np.uint8)
# Overlay the contour on the result image without blending
result_image[solid_boundary_mask > 0] = contour_color # Set contour pixels to blue
# Draw dots at the specified locations
dot_radius = 6
for location, label in zip(dot_locations, dot_labels):
if label:
cv2.circle(result_image, location, dot_radius, COLORS[label], -1)
else:
cv2.drawMarker(result_image, location, COLORS[label], markerType=MARKERS[label],
markerSize=6, thickness=3)
return Image.fromarray(result_image)
def blend_seg_pure(input_image, segmented_image, dot_locations, dot_labels):
input_image = np.array(input_image)
segmented_image = np.array(segmented_image)
# Create a mask for the foreground (non-transparent) pixels
foreground_mask = segmented_image[:, :, 3] > 0
# Blend the foreground
red_foreground = input_image.copy()
blend_weight = 0.8
red_foreground[foreground_mask] = red_foreground[foreground_mask] *(1-blend_weight) + np.array((255,0,0)) * blend_weight
result_image = red_foreground
# Draw dots at the specified locations
dot_radius = 6
for location, label in zip(dot_locations, dot_labels):
if label:
cv2.circle(result_image, location, dot_radius, COLORS[label], -1)
else:
cv2.drawMarker(result_image, location, COLORS[label], markerType=MARKERS[label],
markerSize=6, thickness=3)
return Image.fromarray(result_image)
|