File size: 10,708 Bytes
17f1739
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
import cv2
import numpy as np
from PIL import Image, ImageEnhance
from typing import List, Optional, Tuple
import logging
import asyncio
from pathlib import Path
from pdf2image import convert_from_path
from typing import Dict

logger = logging.getLogger(__name__)

class ImageProcessor:
    """Image processing utilities for certificate analysis"""
    
    def __init__(self):
        self.supported_formats = ['.pdf', '.jpg', '.jpeg', '.png', '.tiff', '.bmp']
        self.model_version = "2024.1.0-image-processor"
        
        logger.info(f"ImageProcessor v{self.model_version} initialized")
    
    async def process_document(self, document_path: str) -> List[np.ndarray]:
        """Process document and extract images from all pages"""
        try:
            images = []
            
            # Check file type
            file_path = Path(document_path)
            suffix = file_path.suffix.lower()
            
            if suffix == '.pdf':
                images = await self._extract_from_pdf(document_path)
            elif suffix in ['.jpg', '.jpeg', '.png', '.tiff', '.bmp']:
                images = await self._extract_from_image(document_path)
            else:
                raise ValueError(f"Unsupported file format: {suffix}")
            
            if not images:
                raise Exception("No images extracted from document")
            
            # Process each image
            processed_images = []
            for img in images:
                processed = await self._process_image(img)
                if processed is not None:
                    processed_images.append(processed)
            
            logger.info(f"Extracted {len(processed_images)} processed images")
            return processed_images
            
        except Exception as e:
            logger.error(f"Document processing failed: {e}")
            raise
    
    async def _extract_from_pdf(self, pdf_path: str) -> List[np.ndarray]:
        """Extract images from PDF"""
        try:
           
            
            images = convert_from_path(pdf_path, dpi=200)
            image_arrays = []
            
            for img in images:
                # Convert PIL to numpy
                img_array = np.array(img)
                
                # Convert RGB to BGR for OpenCV if needed
                if len(img_array.shape) == 3 and img_array.shape[2] == 3:
                    img_array = cv2.cvtColor(img_array, cv2.COLOR_RGB2BGR)
                
                image_arrays.append(img_array)
            
            return image_arrays
            
        except ImportError:
            logger.error("pdf2image not installed. Install with: pip install pdf2image")
            raise
        except Exception as e:
            logger.error(f"PDF extraction failed: {e}")
            raise
    
    async def _extract_from_image(self, image_path: str) -> List[np.ndarray]:
        """Load image from file"""
        try:
            img = cv2.imread(image_path)
            if img is None:
                raise Exception(f"Failed to load image: {image_path}")
            
            return [img]
            
        except Exception as e:
            logger.error(f"Image loading failed: {e}")
            raise
    
    async def _process_image(self, image: np.ndarray) -> Optional[np.ndarray]:
        """Process single image for analysis"""
        try:
            # Convert to RGB if needed
            if len(image.shape) == 3 and image.shape[2] == 4:
                image = cv2.cvtColor(image, cv2.COLOR_BGRA2BGR)
            elif len(image.shape) == 3 and image.shape[2] == 3:
                pass  # Already BGR
            elif len(image.shape) == 2:
                image = cv2.cvtColor(image, cv2.COLOR_GRAY2BGR)
            else:
                raise ValueError(f"Unsupported image shape: {image.shape}")
            
            # Resize if too large
            h, w = image.shape[:2]
            max_dimension = 2000
            
            if max(h, w) > max_dimension:
                scale = max_dimension / max(h, w)
                new_h, new_w = int(h * scale), int(w * scale)
                image = cv2.resize(image, (new_w, new_h), interpolation=cv2.INTER_AREA)
            
            # Enhance image for better OCR
            enhanced = await self._enhance_image(image)
            
            return enhanced
            
        except Exception as e:
            logger.error(f"Image processing failed: {e}")
            return None
    
    async def _enhance_image(self, image: np.ndarray) -> np.ndarray:
        """Enhance image for OCR"""
        try:
            # Convert to PIL for enhancement
            pil_img = Image.fromarray(cv2.cvtColor(image, cv2.COLOR_BGR2RGB))
            
            # Enhance contrast
            enhancer = ImageEnhance.Contrast(pil_img)
            pil_img = enhancer.enhance(1.2)
            
            # Enhance sharpness
            enhancer = ImageEnhance.Sharpness(pil_img)
            pil_img = enhancer.enhance(1.1)
            
            # Convert back to numpy
            enhanced = cv2.cvtColor(np.array(pil_img), cv2.COLOR_RGB2BGR)
            
            # Additional OpenCV enhancements
            # Convert to LAB color space
            lab = cv2.cvtColor(enhanced, cv2.COLOR_BGR2LAB)
            l, a, b = cv2.split(lab)
            
            # Apply CLAHE to L-channel
            clahe = cv2.createCLAHE(clipLimit=2.0, tileGridSize=(8, 8))
            cl = clahe.apply(l)
            
            # Merge channels
            limg = cv2.merge([cl, a, b])
            enhanced = cv2.cvtColor(limg, cv2.COLOR_LAB2BGR)
            
            # Denoise - REMOVED for performance (too slow on CPU)
            # enhanced = cv2.fastNlMeansDenoisingColored(enhanced, None, 10, 10, 7, 21)
            
            return enhanced
            
        except Exception as e:
            logger.debug(f"Image enhancement failed, using original: {e}")
            return image
    
    def deskew_image(self, image: np.ndarray) -> np.ndarray:
        """Deskew image if tilted"""
        try:
            gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
            gray = cv2.bitwise_not(gray)
            
            # Threshold the image
            thresh = cv2.threshold(gray, 0, 255, cv2.THRESH_BINARY | cv2.THRESH_OTSU)[1]
            
            # Find coordinates of non-zero pixels
            coords = np.column_stack(np.where(thresh > 0))
            
            # Get angle of minimum area rectangle
            angle = cv2.minAreaRect(coords)[-1]
            
            # Adjust angle
            if angle < -45:
                angle = 90 + angle
            else:
                angle = -angle
            
            # Rotate image if angle is significant
            if abs(angle) > 0.5:
                (h, w) = image.shape[:2]
                center = (w // 2, h // 2)
                M = cv2.getRotationMatrix2D(center, angle, 1.0)
                rotated = cv2.warpAffine(image, M, (w, h), 
                                        flags=cv2.INTER_CUBIC,
                                        borderMode=cv2.BORDER_REPLICATE)
                return rotated
            
            return image
            
        except Exception as e:
            logger.debug(f"Deskew failed: {e}")
            return image
    
    def detect_and_crop_edges(self, image: np.ndarray) -> np.ndarray:
        """Detect and crop edges to remove borders"""
        try:
            gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
            
            # Apply edge detection
            edges = cv2.Canny(gray, 50, 150)
            
            # Find contours
            contours, _ = cv2.findContours(edges, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
            
            if contours:
                # Find largest contour (assumed to be document)
                largest_contour = max(contours, key=cv2.contourArea)
                x, y, w, h = cv2.boundingRect(largest_contour)
                
                # Add padding
                padding = 20
                x = max(0, x - padding)
                y = max(0, y - padding)
                w = min(image.shape[1] - x, w + 2 * padding)
                h = min(image.shape[0] - y, h + 2 * padding)
                
                # Crop image
                cropped = image[y:y+h, x:x+w]
                
                # Only return if crop is significantly smaller than original
                if w < image.shape[1] * 0.9 or h < image.shape[0] * 0.9:
                    return cropped
            
            return image
            
        except Exception as e:
            logger.debug(f"Edge cropping failed: {e}")
            return image
    
    def calculate_image_quality(self, image: np.ndarray) -> Dict[str, float]:
        """Calculate image quality metrics"""
        try:
            gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
            
            # Calculate blurriness (Laplacian variance)
            blur_score = cv2.Laplacian(gray, cv2.CV_64F).var()
            
            # Calculate contrast
            contrast_score = np.std(gray)
            
            # Calculate brightness
            brightness_score = np.mean(gray)
            
            # Calculate entropy
            hist = cv2.calcHist([gray], [0], None, [256], [0, 256])
            hist = hist.ravel() / hist.sum()
            entropy_score = -np.sum(hist * np.log2(hist + 1e-10))
            
            # Calculate noise
            denoised = cv2.fastNlMeansDenoising(gray, None, 10, 7, 21)
            noise_score = np.mean(np.abs(gray.astype(float) - denoised.astype(float)))
            
            return {
                'blur_score': float(blur_score),
                'contrast_score': float(contrast_score),
                'brightness_score': float(brightness_score),
                'entropy_score': float(entropy_score),
                'noise_score': float(noise_score),
                'overall_quality': float(min(
                    (blur_score / 100) * 0.3 +
                    (contrast_score / 50) * 0.3 +
                    (1 - brightness_score / 255) * 0.2 +
                    (entropy_score / 8) * 0.2,
                    1.0
                ))
            }
            
        except Exception as e:
            logger.debug(f"Quality calculation failed: {e}")
            return {
                'blur_score': 0.0,
                'contrast_score': 0.0,
                'brightness_score': 0.0,
                'entropy_score': 0.0,
                'noise_score': 0.0,
                'overall_quality': 0.0
            }