swoyam-sarvam
commited on
Commit
·
c48064c
1
Parent(s):
cd8f732
added mask creation
Browse files- app.py +18 -10
- mask.py +45 -0
- requirements.txt +4 -0
app.py
CHANGED
@@ -19,28 +19,36 @@ install_packages()
|
|
19 |
import gradio as gr
|
20 |
from control import main
|
21 |
from diffusers.utils import load_image
|
|
|
22 |
|
23 |
-
|
24 |
-
def process_images(image, mask, prompt):
|
25 |
try:
|
26 |
-
#
|
|
|
|
|
|
|
|
|
|
|
|
|
27 |
result = main(image, mask, prompt)
|
28 |
-
|
|
|
29 |
except Exception as e:
|
30 |
-
|
31 |
-
|
32 |
|
33 |
# Create Gradio interface
|
34 |
demo = gr.Interface(
|
35 |
-
fn=
|
36 |
inputs=[
|
37 |
gr.Image(label="Input Image", type="pil"),
|
38 |
-
gr.Image(label="Mask Image", type="pil"),
|
39 |
gr.Textbox(label="Prompt"),
|
40 |
],
|
41 |
-
outputs=
|
|
|
|
|
|
|
42 |
title="Image Inpainting with FLUX ControlNet",
|
43 |
-
description="Upload an image and
|
44 |
)
|
45 |
|
46 |
if __name__ == "__main__":
|
|
|
19 |
import gradio as gr
|
20 |
from control import main
|
21 |
from diffusers.utils import load_image
|
22 |
+
from mask import create_mask
|
23 |
|
24 |
+
def process_image(image, prompt):
|
|
|
25 |
try:
|
26 |
+
# Create mask from input image
|
27 |
+
mask = create_mask(image)
|
28 |
+
|
29 |
+
# First show the generated mask
|
30 |
+
yield mask
|
31 |
+
|
32 |
+
# Then process image with mask
|
33 |
result = main(image, mask, prompt)
|
34 |
+
yield result
|
35 |
+
|
36 |
except Exception as e:
|
37 |
+
yield str(e)
|
|
|
38 |
|
39 |
# Create Gradio interface
|
40 |
demo = gr.Interface(
|
41 |
+
fn=process_image,
|
42 |
inputs=[
|
43 |
gr.Image(label="Input Image", type="pil"),
|
|
|
44 |
gr.Textbox(label="Prompt"),
|
45 |
],
|
46 |
+
outputs=[
|
47 |
+
gr.Image(label="Generated Mask"),
|
48 |
+
gr.Image(label="Generated Image")
|
49 |
+
],
|
50 |
title="Image Inpainting with FLUX ControlNet",
|
51 |
+
description="Upload an image and provide a prompt. The system will first generate a mask and then create the inpainted result.",
|
52 |
)
|
53 |
|
54 |
if __name__ == "__main__":
|
mask.py
ADDED
@@ -0,0 +1,45 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import cv2
|
2 |
+
import numpy as np
|
3 |
+
import matplotlib.pyplot as plt
|
4 |
+
|
5 |
+
def create_mask(image):
|
6 |
+
"""
|
7 |
+
Creates a binary mask from an input image by thresholding and smoothing.
|
8 |
+
|
9 |
+
Args:
|
10 |
+
image: Input image (numpy array)
|
11 |
+
|
12 |
+
Returns:
|
13 |
+
Binary mask image (numpy array)
|
14 |
+
"""
|
15 |
+
# Store original image for visualization
|
16 |
+
image_org = image.copy()
|
17 |
+
|
18 |
+
# Convert image to binary (0 or 255)
|
19 |
+
for i in range(len(image)):
|
20 |
+
for j in range(len(image[i])):
|
21 |
+
if image[i][j] != 255:
|
22 |
+
image[i][j] = 0
|
23 |
+
|
24 |
+
# Add padding of 50 pixels on all sides
|
25 |
+
padding = 0
|
26 |
+
image = cv2.copyMakeBorder(image, padding, padding, padding, padding,
|
27 |
+
cv2.BORDER_CONSTANT, value=255)
|
28 |
+
|
29 |
+
# Apply Gaussian blur for smoothening
|
30 |
+
image = cv2.GaussianBlur(image, (5,5), 50)
|
31 |
+
|
32 |
+
# Threshold to create binary mask
|
33 |
+
_, mask = cv2.threshold(image, 254, 255, cv2.THRESH_BINARY)
|
34 |
+
|
35 |
+
# Visualization (commented out for production use)
|
36 |
+
"""
|
37 |
+
fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(10, 5))
|
38 |
+
ax1.imshow(image_org, cmap="gray")
|
39 |
+
ax1.set_title("Original Image")
|
40 |
+
ax2.imshow(mask, cmap="gray")
|
41 |
+
ax2.set_title("Mask Image")
|
42 |
+
plt.show()
|
43 |
+
"""
|
44 |
+
|
45 |
+
return mask
|
requirements.txt
ADDED
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
1 |
+
diffusers==0.30.2
|
2 |
+
torch>=2.6.0
|
3 |
+
gradio>=5.17.1
|
4 |
+
transformers>=4.49.0
|