File size: 15,101 Bytes
3b3c48f |
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 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 |
"""
Utility functions for the Interior Style Transfer Pipeline
"""
import cv2
import numpy as np
from PIL import Image
import os
from typing import Tuple, List, Optional, Union
import json
from pathlib import Path
def load_image_safe(image_path: str, target_size: Tuple[int, int] = None) -> np.ndarray:
"""
Safely load an image with error handling
Args:
image_path: Path to the image file
target_size: Optional target size (width, height)
Returns:
Loaded image as numpy array
Raises:
ValueError: If image cannot be loaded
"""
if not os.path.exists(image_path):
raise ValueError(f"Image file not found: {image_path}")
# Try to load with OpenCV first
image = cv2.imread(image_path)
if image is None:
# Fallback to PIL
try:
pil_image = Image.open(image_path)
image = np.array(pil_image)
if len(image.shape) == 3 and image.shape[2] == 3:
image = cv2.cvtColor(image, cv2.COLOR_RGB2BGR)
elif len(image.shape) == 3 and image.shape[2] == 4:
image = cv2.cvtColor(image, cv2.COLOR_RGBA2BGR)
except Exception as e:
raise ValueError(f"Could not load image {image_path}: {e}")
if target_size:
image = cv2.resize(image, target_size)
return image
def save_image_safe(image: np.ndarray, output_path: str,
quality: int = 95) -> bool:
"""
Safely save an image with error handling
Args:
image: Image to save as numpy array
output_path: Output file path
quality: JPEG quality (1-100)
Returns:
True if successful, False otherwise
"""
try:
# Ensure output directory exists
os.makedirs(os.path.dirname(output_path), exist_ok=True)
# Save with OpenCV
success = cv2.imwrite(output_path, image)
if not success:
# Fallback to PIL
if len(image.shape) == 3 and image.shape[2] == 3:
pil_image = Image.fromarray(cv2.cvtColor(image, cv2.COLOR_BGR2RGB))
else:
pil_image = Image.fromarray(image)
pil_image.save(output_path, quality=quality)
success = True
return success
except Exception as e:
print(f"Error saving image to {output_path}: {e}")
return False
def validate_image_pair(user_room: np.ndarray, inspiration_room: np.ndarray) -> Tuple[bool, str]:
"""
Validate that two images are suitable for style transfer
Args:
user_room: User room image
inspiration_room: Inspiration room image
Returns:
Tuple of (is_valid, error_message)
"""
# Check image dimensions
if user_room.shape != inspiration_room.shape:
return False, f"Image dimensions don't match: {user_room.shape} vs {inspiration_room.shape}"
# Check minimum size
min_size = 256
if user_room.shape[0] < min_size or user_room.shape[1] < min_size:
return False, f"Images too small. Minimum size: {min_size}x{min_size}"
# Check aspect ratio (should be roughly square for best results)
aspect_ratio = user_room.shape[1] / user_room.shape[0]
if aspect_ratio < 0.5 or aspect_ratio > 2.0:
return False, f"Extreme aspect ratio: {aspect_ratio:.2f}. Square images work best."
# Check if images are too dark or too bright
user_brightness = np.mean(cv2.cvtColor(user_room, cv2.COLOR_BGR2GRAY))
inspiration_brightness = np.mean(cv2.cvtColor(inspiration_room, cv2.COLOR_BGR2GRAY))
if user_brightness < 30 or user_brightness > 225:
return False, f"User room too {'dark' if user_brightness < 30 else 'bright'}: {user_brightness:.1f}"
if inspiration_brightness < 30 or inspiration_brightness > 225:
return False, f"Inspiration room too {'dark' if inspiration_brightness < 30 else 'bright'}: {inspiration_brightness:.1f}"
return True, "Images are valid for style transfer"
def create_comparison_image(original: np.ndarray, result: np.ndarray,
title: str = "Style Transfer Comparison") -> np.ndarray:
"""
Create a side-by-side comparison image
Args:
original: Original user room image
result: Style transfer result
title: Title for the comparison
Returns:
Comparison image
"""
# Ensure both images have the same dimensions
if original.shape != result.shape:
result = cv2.resize(result, (original.shape[1], original.shape[0]))
# Create comparison image
comparison = np.hstack([original, result])
# Add title
font = cv2.FONT_HERSHEY_SIMPLEX
font_scale = 1.0
thickness = 2
# Calculate text position
text_size = cv2.getTextSize(title, font, font_scale, thickness)[0]
text_x = (comparison.shape[1] - text_size[0]) // 2
text_y = 50
# Add background for text
cv2.rectangle(comparison, (text_x - 10, text_y - 30),
(text_x + text_size[0] + 10, text_y + 10), (255, 255, 255), -1)
# Add text
cv2.putText(comparison, title, (text_x, text_y), font, font_scale, (0, 0, 0), thickness)
# Add labels
cv2.putText(comparison, "Original", (50, comparison.shape[0] - 30),
font, 0.7, (255, 255, 255), 2)
cv2.putText(comparison, "Result", (original.shape[1] + 50, comparison.shape[0] - 30),
font, 0.7, (255, 255, 255), 2)
return comparison
def create_multi_comparison_image(images: List[np.ndarray],
titles: List[str] = None,
title: str = "Multi-Image Comparison") -> np.ndarray:
"""
Create a comparison image with multiple images side by side
Args:
images: List of images to compare
titles: List of titles for each image (optional)
title: Main title for the comparison
Returns:
Comparison image
"""
if not images:
raise ValueError("At least one image is required")
# Ensure all images have the same dimensions
target_shape = images[0].shape
resized_images = []
for img in images:
if img.shape != target_shape:
resized_img = cv2.resize(img, (target_shape[1], target_shape[0]))
resized_images.append(resized_img)
else:
resized_images.append(img)
# Create horizontal stack of images
comparison = np.hstack(resized_images)
# Add main title
font = cv2.FONT_HERSHEY_SIMPLEX
font_scale = 1.0
thickness = 2
# Calculate text position for main title
text_size = cv2.getTextSize(title, font, font_scale, thickness)[0]
text_x = (comparison.shape[1] - text_size[0]) // 2
text_y = 50
# Add background for main title
cv2.rectangle(comparison, (text_x - 10, text_y - 30),
(text_x + text_size[0] + 10, text_y + 10), (255, 255, 255), -1)
# Add main title
cv2.putText(comparison, title, (text_x, text_y), font, font_scale, (0, 0, 0), thickness)
# Add individual image titles if provided
if titles and len(titles) == len(images):
font_scale_small = 0.7
thickness_small = 1
for i, (img, img_title) in enumerate(zip(resized_images, titles)):
# Calculate position for each image title
img_width = img.shape[1]
start_x = sum(img.shape[1] for img in resized_images[:i])
# Add background for image title
title_size = cv2.getTextSize(img_title, font, font_scale_small, thickness_small)[0]
title_x = start_x + (img_width - title_size[0]) // 2
title_y = comparison.shape[0] - 30
# Add background rectangle
cv2.rectangle(comparison, (title_x - 5, title_y - 20),
(title_x + title_size[0] + 5, title_y + 5), (255, 255, 255), -1)
# Add image title
cv2.putText(comparison, img_title, (title_x, title_y),
font, font_scale_small, (0, 0, 0), thickness_small)
return comparison
def enhance_image_quality(image: np.ndarray,
sharpness: float = 0.3,
contrast: float = 1.1,
saturation: float = 1.1) -> np.ndarray:
"""
Enhance image quality with various filters
Args:
image: Input image
sharpness: Sharpening strength (0.0 to 1.0)
contrast: Contrast multiplier
saturation: Saturation multiplier
Returns:
Enhanced image
"""
enhanced = image.copy()
# Sharpening
if sharpness > 0:
kernel = np.array([[-1, -1, -1],
[-1, 9, -1],
[-1, -1, -1]]) * sharpness
enhanced = cv2.filter2D(enhanced, -1, kernel)
# Contrast adjustment
if contrast != 1.0:
enhanced = np.clip(enhanced * contrast, 0, 255).astype(np.uint8)
# Saturation adjustment
if saturation != 1.0:
hsv = cv2.cvtColor(enhanced, cv2.COLOR_BGR2HSV).astype(np.float32)
hsv[:, :, 1] = np.clip(hsv[:, :, 1] * saturation, 0, 255)
enhanced = cv2.cvtColor(hsv.astype(np.uint8), cv2.COLOR_HSV2BGR)
return enhanced
def create_progress_bar(total: int, description: str = "Processing") -> callable:
"""
Create a simple progress bar function
Args:
total: Total number of steps
description: Description of the process
Returns:
Function to update progress
"""
def update_progress(current: int):
percentage = (current / total) * 100
bar_length = 30
filled_length = int(bar_length * current // total)
bar = '█' * filled_length + '-' * (bar_length - filled_length)
print(f'\r{description}: |{bar}| {percentage:.1f}% ({current}/{total})', end='')
if current == total:
print()
return update_progress
def save_metadata(metadata: dict, output_path: str) -> bool:
"""
Save metadata to JSON file
Args:
metadata: Dictionary of metadata
output_path: Output file path
Returns:
True if successful, False otherwise
"""
try:
os.makedirs(os.path.dirname(output_path), exist_ok=True)
with open(output_path, 'w') as f:
json.dump(metadata, f, indent=2, default=str)
return True
except Exception as e:
print(f"Error saving metadata to {output_path}: {e}")
return False
def load_metadata(metadata_path: str) -> Optional[dict]:
"""
Load metadata from JSON file
Args:
metadata_path: Path to metadata file
Returns:
Loaded metadata dictionary or None if failed
"""
try:
with open(metadata_path, 'r') as f:
return json.load(f)
except Exception as e:
print(f"Error loading metadata from {metadata_path}: {e}")
return None
def calculate_image_similarity(img1: np.ndarray, img2: np.ndarray) -> float:
"""
Calculate similarity between two images using structural similarity
Args:
img1: First image
img2: Second image
Returns:
Similarity score (0.0 to 1.0, higher is more similar)
"""
try:
from skimage.metrics import structural_similarity as ssim
# Ensure same dimensions
if img1.shape != img2.shape:
img2 = cv2.resize(img2, (img1.shape[1], img1.shape[0]))
# Convert to grayscale for SSIM
gray1 = cv2.cvtColor(img1, cv2.COLOR_BGR2GRAY)
gray2 = cv2.cvtColor(img2, cv2.COLOR_BGR2GRAY)
# Calculate SSIM
similarity = ssim(gray1, gray2)
return max(0.0, similarity) # Ensure non-negative
except ImportError:
# Fallback to simple MSE-based similarity
if img1.shape != img2.shape:
img2 = cv2.resize(img2, (img1.shape[1], img1.shape[0]))
mse = np.mean((img1.astype(np.float32) - img2.astype(np.float32)) ** 2)
max_mse = 255 ** 2
similarity = 1.0 - (mse / max_mse)
return max(0.0, similarity)
def create_thumbnail(image: np.ndarray, max_size: int = 200) -> np.ndarray:
"""
Create a thumbnail version of an image
Args:
image: Input image
max_size: Maximum dimension size
Returns:
Thumbnail image
"""
height, width = image.shape[:2]
if height <= max_size and width <= max_size:
return image.copy()
# Calculate new dimensions maintaining aspect ratio
if height > width:
new_height = max_size
new_width = int(width * max_size / height)
else:
new_width = max_size
new_height = int(height * max_size / width)
thumbnail = cv2.resize(image, (new_width, new_height))
return thumbnail
def batch_resize_images(images: List[np.ndarray],
target_size: Tuple[int, int]) -> List[np.ndarray]:
"""
Resize a list of images to the same target size
Args:
images: List of input images
target_size: Target size (width, height)
Returns:
List of resized images
"""
resized_images = []
for image in images:
resized = cv2.resize(image, target_size)
resized_images.append(resized)
return resized_images
def create_image_grid(images: List[np.ndarray],
grid_size: Tuple[int, int] = None) -> np.ndarray:
"""
Create a grid layout of images
Args:
images: List of images to arrange in grid
grid_size: Grid dimensions (rows, cols). If None, auto-calculate
Returns:
Grid image
"""
if not images:
return np.array([])
if grid_size is None:
# Auto-calculate grid size
n_images = len(images)
cols = int(np.ceil(np.sqrt(n_images)))
rows = int(np.ceil(n_images / cols))
grid_size = (rows, cols)
rows, cols = grid_size
# Ensure all images have the same size
target_size = (images[0].shape[1], images[0].shape[0])
resized_images = batch_resize_images(images, target_size)
# Create grid
grid_rows = []
for i in range(rows):
row_images = []
for j in range(cols):
idx = i * cols + j
if idx < len(resized_images):
row_images.append(resized_images[idx])
else:
# Fill empty space with black
empty_image = np.zeros((target_size[1], target_size[0], 3), dtype=np.uint8)
row_images.append(empty_image)
row = np.hstack(row_images)
grid_rows.append(row)
grid = np.vstack(grid_rows)
return grid
|