Spaces:
Running
Running
File size: 33,721 Bytes
9a6a4dc |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 |
"""
Video Content Analyzer for GAIA Agent - Phase 5
Provides comprehensive video content analysis including scene segmentation, temporal patterns, and content summarization.
Features:
- Scene segmentation and analysis
- Temporal pattern recognition
- Object interaction analysis
- Content summarization and reporting
- Key frame identification and extraction
- Video metadata analysis
"""
import os
import logging
import cv2
import numpy as np
from typing import Dict, Any, List, Optional, Tuple
import json
from datetime import datetime, timedelta
from pathlib import Path
import tempfile
# Configure logging
logger = logging.getLogger(__name__)
class VideoContentAnalyzer:
"""Advanced video content analyzer for scene understanding and temporal analysis."""
def __init__(self):
"""Initialize the video content analyzer."""
self.available = True
self.temp_dir = tempfile.mkdtemp()
# Analysis parameters
self.scene_change_threshold = 0.3
self.keyframe_interval = 30 # Extract keyframe every 30 frames
self.min_scene_duration = 2.0 # Minimum scene duration in seconds
self.max_scenes = 50 # Maximum number of scenes to analyze
# Initialize analysis components
self._init_scene_analyzer()
self._init_temporal_analyzer()
logger.info(f"πΉ Video Content Analyzer initialized - Available: {self.available}")
def _init_scene_analyzer(self):
"""Initialize scene analysis components."""
try:
# Scene change detection parameters
self.scene_detector_params = {
'histogram_bins': 32,
'color_spaces': ['HSV', 'RGB'],
'comparison_methods': [cv2.HISTCMP_CORREL, cv2.HISTCMP_CHISQR],
'motion_threshold': 0.1
}
logger.info("β
Scene analyzer initialized")
except Exception as e:
logger.warning(f"β οΈ Scene analyzer initialization failed: {e}")
def _init_temporal_analyzer(self):
"""Initialize temporal analysis components."""
try:
# Temporal pattern analysis parameters
self.temporal_params = {
'pattern_window': 10, # Analyze patterns over 10 frame windows
'smoothing_factor': 0.3,
'trend_threshold': 0.1,
'periodicity_detection': True
}
logger.info("β
Temporal analyzer initialized")
except Exception as e:
logger.warning(f"β οΈ Temporal analyzer initialization failed: {e}")
def analyze_video_content(self, video_path: str,
object_detections: List[List[Dict[str, Any]]] = None,
question: str = None) -> Dict[str, Any]:
"""
Perform comprehensive video content analysis.
Args:
video_path: Path to video file
object_detections: Optional pre-computed object detections per frame
question: Optional question to guide analysis
Returns:
Comprehensive content analysis results
"""
try:
logger.info(f"πΉ Starting video content analysis for: {video_path}")
# Extract video metadata
metadata = self._extract_video_metadata(video_path)
# Perform scene segmentation
scenes = self._segment_scenes(video_path)
# Extract key frames
keyframes = self._extract_keyframes(video_path, scenes)
# Analyze temporal patterns
temporal_analysis = self._analyze_temporal_patterns(
video_path, object_detections, scenes
)
# Perform content summarization
content_summary = self._summarize_content(
scenes, keyframes, temporal_analysis, object_detections
)
# Generate interaction analysis
interaction_analysis = self._analyze_object_interactions(
object_detections, scenes
)
# Create comprehensive report
analysis_report = self._create_content_report(
metadata, scenes, keyframes, temporal_analysis,
content_summary, interaction_analysis, question
)
return analysis_report
except Exception as e:
logger.error(f"β Video content analysis failed: {e}")
return {
'success': False,
'error': f'Content analysis failed: {str(e)}'
}
def _extract_video_metadata(self, video_path: str) -> Dict[str, Any]:
"""Extract comprehensive video metadata."""
try:
cap = cv2.VideoCapture(video_path)
if not cap.isOpened():
raise Exception("Failed to open video file")
# Basic properties
fps = cap.get(cv2.CAP_PROP_FPS)
frame_count = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))
width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
duration = frame_count / fps if fps > 0 else 0
# Additional properties
fourcc = int(cap.get(cv2.CAP_PROP_FOURCC))
codec = "".join([chr((fourcc >> 8 * i) & 0xFF) for i in range(4)])
cap.release()
metadata = {
'filename': os.path.basename(video_path),
'duration_seconds': duration,
'fps': fps,
'frame_count': frame_count,
'resolution': {'width': width, 'height': height},
'aspect_ratio': width / height if height > 0 else 1.0,
'codec': codec,
'file_size': os.path.getsize(video_path) if os.path.exists(video_path) else 0,
'analysis_timestamp': datetime.now().isoformat()
}
logger.info(f"π Video metadata extracted: {duration:.1f}s, {width}x{height}, {fps:.1f} FPS")
return metadata
except Exception as e:
logger.error(f"β Failed to extract video metadata: {e}")
return {}
def _segment_scenes(self, video_path: str) -> List[Dict[str, Any]]:
"""Segment video into distinct scenes based on visual changes."""
try:
cap = cv2.VideoCapture(video_path)
if not cap.isOpened():
raise Exception("Failed to open video file")
scenes = []
prev_hist = None
scene_start = 0
frame_count = 0
fps = cap.get(cv2.CAP_PROP_FPS)
scene_id = 0
while cap.isOpened():
ret, frame = cap.read()
if not ret:
break
# Calculate histogram for scene change detection
hsv = cv2.cvtColor(frame, cv2.COLOR_BGR2HSV)
hist = cv2.calcHist([hsv], [0, 1, 2], None,
[self.scene_detector_params['histogram_bins']] * 3,
[0, 180, 0, 256, 0, 256])
# Detect scene change
if prev_hist is not None:
correlation = cv2.compareHist(hist, prev_hist, cv2.HISTCMP_CORREL)
if correlation < self.scene_change_threshold:
# Scene change detected
scene_end = frame_count
scene_duration = (scene_end - scene_start) / fps
if scene_duration >= self.min_scene_duration:
scene = {
'id': scene_id,
'start_frame': scene_start,
'end_frame': scene_end,
'start_time': scene_start / fps,
'end_time': scene_end / fps,
'duration': scene_duration,
'frame_count': scene_end - scene_start
}
scenes.append(scene)
scene_id += 1
if len(scenes) >= self.max_scenes:
break
scene_start = frame_count
prev_hist = hist
frame_count += 1
# Add final scene
if scene_start < frame_count:
scene_duration = (frame_count - scene_start) / fps
if scene_duration >= self.min_scene_duration:
scene = {
'id': scene_id,
'start_frame': scene_start,
'end_frame': frame_count,
'start_time': scene_start / fps,
'end_time': frame_count / fps,
'duration': scene_duration,
'frame_count': frame_count - scene_start
}
scenes.append(scene)
cap.release()
logger.info(f"π¬ Scene segmentation complete: {len(scenes)} scenes detected")
return scenes
except Exception as e:
logger.error(f"β Scene segmentation failed: {e}")
return []
def _extract_keyframes(self, video_path: str, scenes: List[Dict[str, Any]]) -> List[Dict[str, Any]]:
"""Extract representative keyframes from video scenes."""
try:
cap = cv2.VideoCapture(video_path)
if not cap.isOpened():
raise Exception("Failed to open video file")
keyframes = []
fps = cap.get(cv2.CAP_PROP_FPS)
for scene in scenes:
# Extract keyframes from each scene
scene_keyframes = []
# Extract keyframe from middle of scene
mid_frame = (scene['start_frame'] + scene['end_frame']) // 2
cap.set(cv2.CAP_PROP_POS_FRAMES, mid_frame)
ret, frame = cap.read()
if ret:
keyframe = {
'scene_id': scene['id'],
'frame_number': mid_frame,
'timestamp': mid_frame / fps,
'type': 'scene_representative',
'frame_data': frame,
'visual_features': self._extract_visual_features(frame)
}
scene_keyframes.append(keyframe)
# Extract additional keyframes for longer scenes
if scene['duration'] > 10: # For scenes longer than 10 seconds
# Extract keyframes at 1/4 and 3/4 points
for fraction in [0.25, 0.75]:
frame_pos = int(scene['start_frame'] +
fraction * (scene['end_frame'] - scene['start_frame']))
cap.set(cv2.CAP_PROP_POS_FRAMES, frame_pos)
ret, frame = cap.read()
if ret:
keyframe = {
'scene_id': scene['id'],
'frame_number': frame_pos,
'timestamp': frame_pos / fps,
'type': 'temporal_sample',
'frame_data': frame,
'visual_features': self._extract_visual_features(frame)
}
scene_keyframes.append(keyframe)
keyframes.extend(scene_keyframes)
cap.release()
logger.info(f"πΌοΈ Keyframe extraction complete: {len(keyframes)} keyframes extracted")
return keyframes
except Exception as e:
logger.error(f"β Keyframe extraction failed: {e}")
return []
def _extract_visual_features(self, frame: np.ndarray) -> Dict[str, Any]:
"""Extract visual features from a frame."""
try:
features = {}
# Color histogram
hsv = cv2.cvtColor(frame, cv2.COLOR_BGR2HSV)
hist_h = cv2.calcHist([hsv], [0], None, [32], [0, 180])
hist_s = cv2.calcHist([hsv], [1], None, [32], [0, 256])
hist_v = cv2.calcHist([hsv], [2], None, [32], [0, 256])
features['color_histogram'] = {
'hue': hist_h.flatten().tolist(),
'saturation': hist_s.flatten().tolist(),
'value': hist_v.flatten().tolist()
}
# Edge density
gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
edges = cv2.Canny(gray, 50, 150)
edge_density = np.sum(edges > 0) / (edges.shape[0] * edges.shape[1])
features['edge_density'] = float(edge_density)
# Brightness and contrast
features['brightness'] = float(np.mean(gray))
features['contrast'] = float(np.std(gray))
# Dominant colors
features['dominant_colors'] = self._get_dominant_colors(frame)
return features
except Exception as e:
logger.error(f"β Visual feature extraction failed: {e}")
return {}
def _get_dominant_colors(self, frame: np.ndarray, k: int = 3) -> List[List[int]]:
"""Extract dominant colors from frame using k-means clustering."""
try:
# Reshape frame to list of pixels
pixels = frame.reshape(-1, 3)
# Use k-means to find dominant colors
from sklearn.cluster import KMeans
kmeans = KMeans(n_clusters=k, random_state=42, n_init=10)
kmeans.fit(pixels)
# Get dominant colors
colors = kmeans.cluster_centers_.astype(int)
return colors.tolist()
except ImportError:
# Fallback without sklearn
return [[128, 128, 128]] # Gray as default
except Exception as e:
logger.error(f"β Dominant color extraction failed: {e}")
return [[128, 128, 128]]
def _analyze_temporal_patterns(self, video_path: str,
object_detections: List[List[Dict[str, Any]]] = None,
scenes: List[Dict[str, Any]] = None) -> Dict[str, Any]:
"""Analyze temporal patterns in video content."""
try:
temporal_analysis = {
'motion_patterns': [],
'object_appearance_patterns': [],
'scene_transition_patterns': [],
'activity_levels': [],
'periodicity': {}
}
if not object_detections:
return temporal_analysis
# Analyze motion patterns
motion_levels = []
for frame_detections in object_detections:
# Calculate motion level based on number and size of objects
motion_level = len(frame_detections)
if frame_detections:
avg_area = np.mean([det.get('area', 0) for det in frame_detections])
motion_level += avg_area / 10000 # Normalize area contribution
motion_levels.append(motion_level)
temporal_analysis['motion_patterns'] = motion_levels
# Analyze object appearance patterns
object_counts_over_time = []
bird_counts_over_time = []
animal_counts_over_time = []
for frame_detections in object_detections:
object_count = len(frame_detections)
bird_count = sum(1 for det in frame_detections
if det.get('species_type') == 'bird')
animal_count = sum(1 for det in frame_detections
if det.get('species_type') == 'animal')
object_counts_over_time.append(object_count)
bird_counts_over_time.append(bird_count)
animal_counts_over_time.append(animal_count)
temporal_analysis['object_appearance_patterns'] = {
'total_objects': object_counts_over_time,
'birds': bird_counts_over_time,
'animals': animal_counts_over_time
}
# Analyze activity levels
window_size = self.temporal_params['pattern_window']
activity_levels = []
for i in range(0, len(motion_levels), window_size):
window = motion_levels[i:i+window_size]
if window:
activity_level = {
'start_frame': i,
'end_frame': min(i + window_size, len(motion_levels)),
'avg_motion': np.mean(window),
'max_motion': np.max(window),
'motion_variance': np.var(window)
}
activity_levels.append(activity_level)
temporal_analysis['activity_levels'] = activity_levels
# Detect periodicity in object appearances
if len(bird_counts_over_time) > 20: # Need sufficient data
temporal_analysis['periodicity'] = self._detect_periodicity(
bird_counts_over_time, animal_counts_over_time
)
logger.info("π Temporal pattern analysis complete")
return temporal_analysis
except Exception as e:
logger.error(f"β Temporal pattern analysis failed: {e}")
return {}
def _detect_periodicity(self, bird_counts: List[int],
animal_counts: List[int]) -> Dict[str, Any]:
"""Detect periodic patterns in object appearances."""
try:
periodicity = {
'bird_patterns': {},
'animal_patterns': {},
'combined_patterns': {}
}
# Simple autocorrelation-based periodicity detection
def autocorrelation(signal, max_lag=50):
signal = np.array(signal)
n = len(signal)
signal = signal - np.mean(signal)
autocorr = []
for lag in range(min(max_lag, n//2)):
if n - lag > 0:
corr = np.corrcoef(signal[:-lag], signal[lag:])[0, 1]
autocorr.append(corr if not np.isnan(corr) else 0)
else:
autocorr.append(0)
return autocorr
# Analyze bird count periodicity
bird_autocorr = autocorrelation(bird_counts)
if bird_autocorr:
max_corr_idx = np.argmax(bird_autocorr[1:]) + 1 # Skip lag 0
periodicity['bird_patterns'] = {
'dominant_period': max_corr_idx,
'correlation_strength': bird_autocorr[max_corr_idx],
'is_periodic': bird_autocorr[max_corr_idx] > 0.3
}
# Analyze animal count periodicity
animal_autocorr = autocorrelation(animal_counts)
if animal_autocorr:
max_corr_idx = np.argmax(animal_autocorr[1:]) + 1
periodicity['animal_patterns'] = {
'dominant_period': max_corr_idx,
'correlation_strength': animal_autocorr[max_corr_idx],
'is_periodic': animal_autocorr[max_corr_idx] > 0.3
}
return periodicity
except Exception as e:
logger.error(f"β Periodicity detection failed: {e}")
return {}
def _summarize_content(self, scenes: List[Dict[str, Any]],
keyframes: List[Dict[str, Any]],
temporal_analysis: Dict[str, Any],
object_detections: List[List[Dict[str, Any]]] = None) -> Dict[str, Any]:
"""Generate comprehensive content summary."""
try:
summary = {
'overview': {},
'scene_summary': [],
'key_moments': [],
'content_highlights': [],
'statistical_summary': {}
}
# Overview
total_duration = sum(scene.get('duration', 0) for scene in scenes)
summary['overview'] = {
'total_scenes': len(scenes),
'total_duration': total_duration,
'avg_scene_duration': total_duration / len(scenes) if scenes else 0,
'keyframes_extracted': len(keyframes)
}
# Scene summary
for scene in scenes:
scene_summary = {
'scene_id': scene['id'],
'duration': scene['duration'],
'description': f"Scene {scene['id'] + 1}: {scene['duration']:.1f}s",
'activity_level': 'unknown'
}
# Determine activity level from temporal analysis
if temporal_analysis.get('activity_levels'):
scene_start_frame = scene['start_frame']
scene_end_frame = scene['end_frame']
relevant_activities = [
activity for activity in temporal_analysis['activity_levels']
if (activity['start_frame'] <= scene_end_frame and
activity['end_frame'] >= scene_start_frame)
]
if relevant_activities:
avg_motion = np.mean([a['avg_motion'] for a in relevant_activities])
if avg_motion > 2:
scene_summary['activity_level'] = 'high'
elif avg_motion > 1:
scene_summary['activity_level'] = 'medium'
else:
scene_summary['activity_level'] = 'low'
summary['scene_summary'].append(scene_summary)
# Key moments (high activity periods)
if temporal_analysis.get('activity_levels'):
high_activity_moments = [
activity for activity in temporal_analysis['activity_levels']
if activity['avg_motion'] > 2
]
summary['key_moments'] = [
{
'timestamp': moment['start_frame'] / 30, # Assume 30 FPS
'duration': (moment['end_frame'] - moment['start_frame']) / 30,
'activity_level': moment['avg_motion'],
'description': f"High activity period: {moment['avg_motion']:.1f}"
}
for moment in high_activity_moments[:5] # Top 5 moments
]
# Statistical summary
if object_detections:
all_detections = [det for frame_dets in object_detections for det in frame_dets]
species_counts = {}
for detection in all_detections:
species = detection.get('species_type', 'unknown')
species_counts[species] = species_counts.get(species, 0) + 1
summary['statistical_summary'] = {
'total_detections': len(all_detections),
'species_distribution': species_counts,
'avg_detections_per_frame': len(all_detections) / len(object_detections) if object_detections else 0
}
logger.info("π Content summarization complete")
return summary
except Exception as e:
logger.error(f"β Content summarization failed: {e}")
return {}
def _analyze_object_interactions(self, object_detections: List[List[Dict[str, Any]]] = None,
scenes: List[Dict[str, Any]] = None) -> Dict[str, Any]:
"""Analyze interactions between detected objects."""
try:
interaction_analysis = {
'proximity_interactions': [],
'temporal_interactions': [],
'species_interactions': {},
'interaction_summary': {}
}
if not object_detections:
return interaction_analysis
# Analyze proximity interactions within frames
for frame_idx, frame_detections in enumerate(object_detections):
if len(frame_detections) > 1:
# Check all pairs of objects in the frame
for i, obj1 in enumerate(frame_detections):
for j, obj2 in enumerate(frame_detections[i+1:], i+1):
distance = self._calculate_object_distance(obj1, obj2)
if distance < 100: # Close proximity threshold
interaction = {
'frame': frame_idx,
'timestamp': frame_idx / 30, # Assume 30 FPS
'object1': obj1.get('class', 'unknown'),
'object2': obj2.get('class', 'unknown'),
'distance': distance,
'interaction_type': 'proximity'
}
interaction_analysis['proximity_interactions'].append(interaction)
# Analyze species interactions
species_pairs = {}
for interaction in interaction_analysis['proximity_interactions']:
obj1_type = interaction['object1']
obj2_type = interaction['object2']
pair_key = tuple(sorted([obj1_type, obj2_type]))
if pair_key not in species_pairs:
species_pairs[pair_key] = []
species_pairs[pair_key].append(interaction)
interaction_analysis['species_interactions'] = {
f"{pair[0]}-{pair[1]}": {
'interaction_count': len(interactions),
'avg_distance': np.mean([i['distance'] for i in interactions]),
'duration': len(interactions) / 30 # Approximate duration
}
for pair, interactions in species_pairs.items()
}
# Interaction summary
interaction_analysis['interaction_summary'] = {
'total_proximity_interactions': len(interaction_analysis['proximity_interactions']),
'unique_species_pairs': len(species_pairs),
'most_interactive_pair': max(species_pairs.keys(),
key=lambda x: len(species_pairs[x])) if species_pairs else None
}
logger.info("π€ Object interaction analysis complete")
return interaction_analysis
except Exception as e:
logger.error(f"β Object interaction analysis failed: {e}")
return {}
def _calculate_object_distance(self, obj1: Dict[str, Any], obj2: Dict[str, Any]) -> float:
"""Calculate distance between two objects based on their centers."""
try:
center1 = obj1.get('center', [0, 0])
center2 = obj2.get('center', [0, 0])
distance = np.sqrt((center1[0] - center2[0])**2 + (center1[1] - center2[1])**2)
return float(distance)
except Exception as e:
logger.error(f"β Distance calculation failed: {e}")
return float('inf')
def _create_content_report(self, metadata: Dict[str, Any],
scenes: List[Dict[str, Any]],
keyframes: List[Dict[str, Any]],
temporal_analysis: Dict[str, Any],
content_summary: Dict[str, Any],
interaction_analysis: Dict[str, Any],
question: str = None) -> Dict[str, Any]:
"""Create comprehensive content analysis report."""
try:
report = {
'success': True,
'analysis_timestamp': datetime.now().isoformat(),
'question': question,
'metadata': metadata,
'content_analysis': {
'scenes': scenes,
'keyframes': [
{k: v for k, v in kf.items() if k != 'frame_data'} # Exclude frame data
for kf in keyframes
],
'temporal_patterns': temporal_analysis,
'content_summary': content_summary,
'interactions': interaction_analysis
},
'insights': [],
'recommendations': []
}
# Generate insights
insights = []
# Scene insights
if scenes:
avg_scene_duration = np.mean([s['duration'] for s in scenes])
insights.append(f"Video contains {len(scenes)} distinct scenes with average duration of {avg_scene_duration:.1f}s")
# Activity insights
if temporal_analysis.get('activity_levels'):
high_activity_count = sum(1 for a in temporal_analysis['activity_levels'] if a['avg_motion'] > 2)
insights.append(f"Detected {high_activity_count} high-activity periods in the video")
# Interaction insights
if interaction_analysis.get('interaction_summary', {}).get('total_proximity_interactions', 0) > 0:
total_interactions = interaction_analysis['interaction_summary']['total_proximity_interactions']
insights.append(f"Found {total_interactions} object proximity interactions")
report['insights'] = insights
# Generate recommendations
recommendations = []
if question and 'bird' in question.lower():
if temporal_analysis.get('object_appearance_patterns', {}).get('birds'):
max_birds = max(temporal_analysis['object_appearance_patterns']['birds'])
recommendations.append(f"Maximum simultaneous birds detected: {max_birds}")
if len(scenes) > 10:
recommendations.append("Video has many scene changes - consider analyzing key scenes only")
report['recommendations'] = recommendations
logger.info("π Content analysis report generated successfully")
return report
except Exception as e:
logger.error(f"β Failed to create content report: {e}")
return {
'success': False,
'error': f'Failed to create content report: {str(e)}'
}
def get_capabilities(self) -> Dict[str, Any]:
"""Get video content analyzer capabilities."""
return {
'available': self.available,
'scene_change_threshold': self.scene_change_threshold,
'keyframe_interval': self.keyframe_interval,
'min_scene_duration': self.min_scene_duration,
'max_scenes': self.max_scenes,
'features': [
'Scene segmentation',
'Keyframe extraction',
'Temporal pattern analysis',
'Object interaction analysis',
'Content summarization',
'Visual feature extraction',
'Activity level detection',
'Periodicity detection'
]
}
# Factory function for creating content analyzer
def create_video_content_analyzer() -> VideoContentAnalyzer:
"""Create and return a video content analyzer instance."""
return VideoContentAnalyzer()
if __name__ == "__main__":
# Test the content analyzer
analyzer = VideoContentAnalyzer()
print(f"Content analyzer available: {analyzer.available}")
print(f"Capabilities: {json.dumps(analyzer.get_capabilities(), indent=2)}") |