|
|
import gradio as gr |
|
|
import numpy as np |
|
|
from io import BytesIO |
|
|
from PIL import Image, ImageOps |
|
|
import zipfile |
|
|
import os |
|
|
import atexit |
|
|
import shutil |
|
|
import cv2 |
|
|
import imageio |
|
|
import torchvision.transforms.functional as TF |
|
|
|
|
|
|
|
|
GENERATED_FILES_DIR = "generated_files" |
|
|
if not os.path.exists(GENERATED_FILES_DIR): |
|
|
os.makedirs(GENERATED_FILES_DIR) |
|
|
|
|
|
def cleanup_generated_files(): |
|
|
if os.path.exists(GENERATED_FILES_DIR): |
|
|
shutil.rmtree(GENERATED_FILES_DIR) |
|
|
|
|
|
|
|
|
atexit.register(cleanup_generated_files) |
|
|
|
|
|
def split_image_grid(image, grid_cols, grid_rows): |
|
|
if isinstance(image, np.ndarray): |
|
|
image = Image.fromarray(image) |
|
|
|
|
|
width, height = image.width, image.height |
|
|
cell_width = width // grid_cols |
|
|
cell_height = height // grid_rows |
|
|
frames = [] |
|
|
for i in range(grid_rows): |
|
|
for j in range(grid_cols): |
|
|
left = j * cell_width |
|
|
upper = i * cell_height |
|
|
right = left + cell_width |
|
|
lower = upper + cell_height |
|
|
frame = image.crop((left, upper, right, lower)) |
|
|
frames.append(np.array(frame)) |
|
|
return frames |
|
|
|
|
|
def interpolate_frames(frames, factor=2): |
|
|
interpolated_frames = [] |
|
|
for i in range(len(frames) - 1): |
|
|
frame1 = frames[i] |
|
|
frame2 = frames[i + 1] |
|
|
interpolated_frames.append(frame1) |
|
|
for j in range(1, factor): |
|
|
t = j / factor |
|
|
frame_t = cv2.addWeighted(frame1, 1 - t, frame2, t, 0) |
|
|
interpolated_frames.append(frame_t) |
|
|
interpolated_frames.append(frames[-1]) |
|
|
return interpolated_frames |
|
|
|
|
|
def enhance_gif(images): |
|
|
enhanced_images = [] |
|
|
for img in images: |
|
|
img = ImageOps.autocontrast(Image.fromarray(img)) |
|
|
img = img.convert("RGB") |
|
|
enhanced_images.append(np.array(img)) |
|
|
return enhanced_images |
|
|
|
|
|
def create_gif_imageio(images, duration=50, loop=0): |
|
|
gif_path = os.path.join(GENERATED_FILES_DIR, "output_enhanced.gif") |
|
|
images_pil = [Image.fromarray(img) for img in images] |
|
|
imageio.mimsave(gif_path, images_pil, duration=duration, loop=loop) |
|
|
return gif_path |
|
|
|
|
|
def process_image(image, grid_cols_input, grid_rows_input): |
|
|
frames = split_image_grid(image, grid_cols_input, grid_rows_input) |
|
|
zip_file = zip_images(frames) |
|
|
return zip_file |
|
|
|
|
|
def process_image_to_gif(image, grid_cols_input, grid_rows_input): |
|
|
frames = split_image_grid(image, grid_cols_input, grid_rows_input) |
|
|
interpolated_frames = interpolate_frames(frames, factor=2) |
|
|
enhanced_frames = enhance_gif(interpolated_frames) |
|
|
gif_file = create_gif_imageio(enhanced_frames, duration=50, loop=0) |
|
|
return gif_file |
|
|
|
|
|
def zip_images(images): |
|
|
zip_path = os.path.join(GENERATED_FILES_DIR, "output.zip") |
|
|
with zipfile.ZipFile(zip_path, 'w') as zipf: |
|
|
for idx, img in enumerate(images): |
|
|
img_buffer = BytesIO() |
|
|
img = Image.fromarray(img) |
|
|
img.save(img_buffer, format='PNG') |
|
|
img_buffer.seek(0) |
|
|
zipf.writestr(f'image_{idx}.png', img_buffer.getvalue()) |
|
|
return zip_path |
|
|
|
|
|
with gr.Blocks() as demo: |
|
|
with gr.Row(): |
|
|
image_input = gr.Image(label="Input Image", type="pil") |
|
|
grid_cols_input = gr.Slider(1, 10, value=2, step=1, label="Grid Columns") |
|
|
grid_rows_input = gr.Slider(1, 10, value=2, step=1, label="Grid Rows") |
|
|
with gr.Row(): |
|
|
zip_button = gr.Button("Create Zip File") |
|
|
gif_button = gr.Button("Create GIF") |
|
|
with gr.Row(): |
|
|
zip_output = gr.File(label="Download Zip File") |
|
|
gif_output = gr.File(label="Download GIF") |
|
|
zip_button.click(process_image, inputs=[image_input, grid_cols_input, grid_rows_input], outputs=zip_output) |
|
|
gif_button.click(process_image_to_gif, inputs=[image_input, grid_cols_input, grid_rows_input], outputs=gif_output) |
|
|
|
|
|
demo.launch(show_error=True) |
