File size: 31,771 Bytes
7baad7b | 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 | # =============================================================================
# MBFS Sentinel - Configuration Template
# =============================================================================
#
# Usage:
# 1. Copy this file to config.yml: cp config.example.yml config.yml
# 2. Update values to match your environment
# 3. Restart mbfs-sentinel to apply changes
#
# Notes:
# - All paths can be relative (to the executable) or absolute
# - Durations use seconds unless noted otherwise
# - Set 'enabled: true' on any plugin you want to activate
#
# =============================================================================
# =============================================================================
# General Settings
# =============================================================================
general:
project_name: MBFS Sentinel # Display name shown in UI and logs
port: 8003 # HTTP API port (Axum server)
dev_mode: false # Enable development mode (extra logging, relaxed checks)
device_id: c9869e48ab4a2341fe4e08587b20d6c402da59ce2e26d934b9d3d6d503e0bd5a # Unique device identifier (from mbfs-mv-core)
# =============================================================================
# Database (PostgreSQL + pgvector)
# =============================================================================
database:
# PostgreSQL connection URL
# Format: postgresql://user:password@host:port/database?sslmode=disable
url: "postgresql://root:123123@localhost:5434/dev?sslmode=disable"
# Secret key for signing JWT tokens (MUST change in production!)
secret_key: "secret"
# =============================================================================
# MinIO Storage (S3-compatible object storage)
# =============================================================================
# Used for storing face crops, event snapshots, evidence images/videos.
# Any S3-compatible service can be used (AWS S3, MinIO, etc.)
# =============================================================================
minio:
endpoint: "localhost:9000" # MinIO server host:port
access_key: "minioadmin" # Access key (username)
secret_key: "minioadmin" # Secret key (password)
bucket: "mbfs-sentinel" # Bucket name for all stored objects
secure: false # Use TLS/SSL connection (true for HTTPS)
url: http://localhost:9000 # Public-facing URL for generating download links
# =============================================================================
# AI Models
# =============================================================================
# Directory and filenames for core ONNX models (face detection & recognition).
# Plugin-specific models are configured in their respective sections below.
# =============================================================================
models:
dir: './models' # Path to model files directory
detection_name: "" # Face detection model (RetinaFace)
recognition_name: "" # Face recognition/embedding model (ArcFace)
# =============================================================================
# Batch Processing
# =============================================================================
# Controls how frames are batched before inference to maximize GPU throughput.
# Larger batches = higher throughput but more latency and VRAM usage.
# =============================================================================
batch:
detection_size: 24 # Face detection batch size (frames per batch)
recognition_size: 96 # Face recognition batch size (face crops per batch)
collection_timeout_ms: 100 # Max wait time to fill a batch before processing (ms)
# =============================================================================
# Worker Threads
# =============================================================================
workers:
num_ai: 3 # Number of AI inference worker threads
upload: 24 # Number of upload worker threads (MinIO/S3)
# =============================================================================
# Face Tracking
# =============================================================================
# Settings for tracking detected faces across consecutive frames.
# Used by face recognition to maintain identity continuity.
# =============================================================================
tracking:
high_confidence_threshold: 0.65 # Min confidence for a high-quality face match
timeout: 2.0 # Seconds before a tracked face expires (no re-detection)
similarity: 0.55 # Cosine similarity threshold for re-identification
log_cooldown: 3.0 # Min seconds between logging the same face identity
min_detections: 2 # Min detection count before a face is logged as event
# =============================================================================
# Image Upload
# =============================================================================
upload:
queue_max_size: 100 # Max pending uploads in queue (overflow is dropped)
jpeg_quality: 70 # JPEG compression quality for uploaded images (0-100)
# =============================================================================
# Camera Manager
# =============================================================================
# Connection to the external camera management service.
# Sentinel periodically sends a survival signal to indicate it is alive.
# =============================================================================
camera_manager:
url: http://localhost:8000 # Camera manager API base URL
survival_signal_interval: 10 # Interval in seconds for survival heartbeat
# =============================================================================
# Core AI Node (Cluster mode)
# =============================================================================
# URL of the central AI node when running in distributed/cluster mode.
# =============================================================================
core_ai:
url: http://10.8.0.3:8003
# =============================================================================
# TensorRT Configuration
# =============================================================================
# TensorRT engine caching speeds up model loading after first run.
# Engine files are GPU-specific and will be rebuilt if hardware changes.
# =============================================================================
tensorrt:
enabled: true # false = skip TensorRT, use CUDA (faster startup, slower inference)
cache_dir: trt_cache # Directory for cached TensorRT engine files
lib_dir: null # Custom TensorRT library path (null = system default)
# =============================================================================
# License
# =============================================================================
license:
key: '' # License key (obtained from license server)
active_key: '' # Active license key (set after successful activation)
license_server:
url: https://ai-mv-core.mbfs.com.vn # License server URL
check_interval_secs: 86400 # License recheck interval (86400 = 24 hours)
# =============================================================================
# Processing Loop
# =============================================================================
# Low-level tuning for the frame processing pipeline.
# Default values work well for most setups. Only adjust if needed.
# =============================================================================
processing:
max_batch_size: 24 # Max frames per processing batch
loop_sleep_ms: 10 # Sleep between processing loops (ms)
channel_buffer_size: 256 # Channel buffer for frame pipeline
face_log_queue_size: 256 # Face event logging queue size
face_log_workers: 2 # Face event logging worker threads
dispatch_max_fps: 15 # Default max dispatch rate per pipeline runner (fps).
# Coordinator drops batches dispatched faster than this
# to avoid wasting Arc clones + mutex contention on frames
# the runner can't keep up with. Camera decode rate is
# capped separately at 15fps in mbfs-camera.
# Plugins can override per-plugin via manifest.toml:
# [runtime]
# max_fps = 8
# =============================================================================
# Plugins - AI Pipeline Configurations
# =============================================================================
#
# Each plugin can be independently enabled/disabled. Common sub-sections:
#
# enabled - Whether the plugin is active (true/false)
# confidence - Min detection confidence threshold (0.0 - 1.0)
# nms_threshold - Non-Maximum Suppression IoU threshold (0.0 - 1.0)
#
# tracking: - ByteTrack multi-object tracker settings
# enabled - Enable/disable object tracking
# high_conf_threshold - High confidence detection threshold
# low_conf_threshold - Low confidence detection threshold
# confirm_frames - Frames needed to confirm a new track
# max_lost_frames - Max frames before a lost track is removed
#
# alert: - Alert system settings (consecutive detection triggers)
# consecutive_threshold - Consecutive detections needed to trigger alert
# max_miss - Max missed frames before resetting alert counter
# cooldown_secs - Cooldown between alerts for the same object
#
# data_lake: - Event storage configuration
# enabled - Enable event logging
# bucket_prefix - MinIO bucket prefix for this plugin's events
# min_count - Min detection count before saving event
# jpeg_quality - JPEG quality for event snapshots (0-100)
# queue_size - Event processing queue size
# workers - Number of event processing workers
# batch_size - Events per batch write
# parquet_enabled - Enable Parquet file export
# parquet_batch_size - Rows per Parquet batch
# save_json - Save raw JSON alongside Parquet
# db_enabled - Write events to PostgreSQL
# db_batch_size - DB insert batch size
# db_batch_timeout_ms - Max wait before flushing DB batch (ms)
# db_channel_size - DB write channel buffer size
# save_crops - Save cropped detection images to MinIO
# crop_jpeg_quality - JPEG quality for crop images (0-100)
#
# dedup: - Deduplication settings (prevent duplicate events)
# cooldown_seconds - Min seconds between events for same object
# grid_size - Spatial grid cell size for position-based dedup
# high_confidence_threshold - Confidence above which dedup is stricter
# significant_count_threshold - Detection count to consider object significant
# max_tracked_per_camera - Max tracked objects per camera for dedup
#
# =============================================================================
plugins:
# ---------------------------------------------------------------------------
# Shared Models
# ---------------------------------------------------------------------------
# Model files referenced by multiple plugins. Avoids duplicating filenames.
# These are loaded once and shared across plugins that need them.
# ---------------------------------------------------------------------------
shared_models:
detection_model: det_10g_fp16_dynamic.onnx.enc # Face detection (RetinaFace)
recognition_model: mbfs_rec_vit_b_v3.onnx.enc # Face recognition (ViT)
human_detection_model: mbfs_human_det_v4.onnx # Human/person detection (YOLO)
helmet_model: helmet_dt_v10.onnx # Helmet detection
lpr_detection_model: lp_detection_v4.onnx # License plate detection
lpr_ocr_model: mbfs_ocr_license_plate_b16_v5.onnx # License plate OCR
vehicle_model: yolov8_coco.onnx # Vehicle detection (COCO classes)
# ---------------------------------------------------------------------------
# Face Recognition (builtin)
# ---------------------------------------------------------------------------
# Detects faces, extracts embeddings, and matches against enrolled identities.
# Requires: detection_model + recognition_model from shared_models.
# ---------------------------------------------------------------------------
face_recognition:
enabled: false
min_detection_score: 0.7 # Min face detection confidence
min_size: 32 # Min face size in pixels (width or height)
min_ratio: 0.0001 # Min face area ratio relative to frame
unknown_min_score: 0.50 # Min score to classify as "unknown" (below = discard)
enrollment_min_detection_score: 0.7 # Stricter detection threshold for face enrollment
enrollment_min_size: 100 # Min face size for enrollment (px)
enrollment_max_angle: 15.0 # Max yaw/pitch angle for enrollment (degrees)
data_lake:
enabled: true
bucket_prefix: face-recognition-events
min_count: 1
jpeg_quality: 85
queue_size: 512
workers: 1
batch_size: 16
parquet_enabled: true
parquet_batch_size: 1000
save_json: false
db_enabled: true
db_batch_size: 32
db_batch_timeout_ms: 100
db_channel_size: 512
save_crops: true
crop_jpeg_quality: 90
dedup:
cooldown_seconds: 5.0
grid_size: 100.0
high_confidence_threshold: 0.65
significant_count_threshold: 3
max_tracked_per_camera: 50
# ===========================================================================
# Dynamic Plugins
# ===========================================================================
# Plugins loaded at runtime from DLLs in the plugins/ directory.
# Each plugin has a manifest.toml defining its models and capabilities.
# Config values here override manifest defaults.
# ===========================================================================
dynamic_plugins:
# -------------------------------------------------------------------------
# Human Detection
# -------------------------------------------------------------------------
# Detects humans/persons in the frame. Foundation for many other plugins.
# Model: human_detection_model from shared_models.
# -------------------------------------------------------------------------
human_detection:
enabled: false
confidence: 0.75
nms_threshold: 0.45
alert:
consecutive_threshold: 5
max_miss: 2
cooldown_secs: 10.0
tracking:
enabled: true
high_conf_threshold: 0.5
low_conf_threshold: 0.1
confirm_frames: 3
max_lost_frames: 30
data_lake:
enabled: true
bucket_prefix: human-detection-events
min_count: 1
jpeg_quality: 85
queue_size: 512
workers: 2
batch_size: 16
parquet_enabled: true
parquet_batch_size: 1000
save_json: false
db_enabled: true
db_batch_size: 16
db_batch_timeout_ms: 100
db_channel_size: 256
save_crops: true
crop_jpeg_quality: 90
dedup:
cooldown_seconds: 3.0
grid_size: 100.0
high_confidence_threshold: 0.75
significant_count_threshold: 5
max_tracked_per_camera: 50
# -------------------------------------------------------------------------
# Facial Expression Recognition
# -------------------------------------------------------------------------
# Classifies facial expressions (happy, sad, angry, surprised, etc.).
# Two-stage: face detection -> expression classification.
# -------------------------------------------------------------------------
facial_expression:
enabled: false
confidence: 0.5
nms_threshold: 0.45
alert:
consecutive_threshold: 5
max_miss: 2
cooldown_secs: 10.0
tracking:
enabled: true
high_conf_threshold: 0.5
low_conf_threshold: 0.1
confirm_frames: 3
max_lost_frames: 30
data_lake:
enabled: true
bucket_prefix: facial-expression-events
min_count: 1
jpeg_quality: 85
queue_size: 512
workers: 2
batch_size: 16
parquet_enabled: true
parquet_batch_size: 1000
save_json: false
db_enabled: true
db_batch_size: 16
db_batch_timeout_ms: 100
db_channel_size: 256
save_crops: true
crop_jpeg_quality: 90
dedup:
cooldown_seconds: 3.0
grid_size: 100.0
high_confidence_threshold: 0.75
significant_count_threshold: 5
max_tracked_per_camera: 50
# -------------------------------------------------------------------------
# Object Detection (General)
# -------------------------------------------------------------------------
# General-purpose object detection using COCO or custom classes.
# Useful for counting/tracking arbitrary object types.
# -------------------------------------------------------------------------
object_detection:
enabled: false
confidence: 0.75
nms_threshold: 0.45
tracking:
enabled: true
high_conf_threshold: 0.5
low_conf_threshold: 0.1
confirm_frames: 3
max_lost_frames: 30
data_lake:
enabled: true
bucket_prefix: object-detection-events
min_count: 1
jpeg_quality: 85
queue_size: 512
workers: 2
batch_size: 16
parquet_enabled: true
parquet_batch_size: 1000
save_json: false
db_enabled: true
db_batch_size: 16
db_batch_timeout_ms: 100
db_channel_size: 256
save_crops: true
crop_jpeg_quality: 90
dedup:
cooldown_seconds: 3.0
grid_size: 100.0
high_confidence_threshold: 0.75
significant_count_threshold: 5
max_tracked_per_camera: 50
# -------------------------------------------------------------------------
# Smoke & Fire Detection
# -------------------------------------------------------------------------
# Detects smoke and fire in the frame. Triggers alerts when detected
# consecutively to reduce false positives.
# -------------------------------------------------------------------------
smoke_fire_detection:
enabled: false
confidence: 0.75
alert:
consecutive_threshold: 5 # Consecutive detections to trigger alert
max_miss: 2 # Max missed frames before reset
cooldown_secs: 10.0 # Alert cooldown (seconds)
data_lake:
enabled: true
bucket_prefix: smoke-fire-detection-events
min_count: 1
jpeg_quality: 85
queue_size: 512
workers: 2
batch_size: 16
parquet_enabled: true
parquet_batch_size: 1000
save_json: false
db_enabled: true
db_batch_size: 16
db_batch_timeout_ms: 100
db_channel_size: 256
save_crops: true
crop_jpeg_quality: 90
dedup:
cooldown_seconds: 3.0
grid_size: 100.0
high_confidence_threshold: 0.75
significant_count_threshold: 5
max_tracked_per_camera: 50
# -------------------------------------------------------------------------
# Behavior Detection
# -------------------------------------------------------------------------
# Detects abnormal behaviors (loitering, falling, running, etc.).
# Uses human detection + tracking to analyze movement patterns.
# -------------------------------------------------------------------------
behavior_detection:
enabled: false
confidence: 0.75
tracking:
enabled: true
high_conf_threshold: 0.5
low_conf_threshold: 0.1
confirm_frames: 5 # Higher confirm frames for behavior analysis
max_lost_frames: 30
data_lake:
enabled: true
bucket_prefix: behavior-detection-events
min_count: 1
jpeg_quality: 85
queue_size: 512
workers: 2
batch_size: 16
parquet_enabled: true
parquet_batch_size: 1000
save_json: false
db_enabled: true
db_batch_size: 16
db_batch_timeout_ms: 100
db_channel_size: 256
save_crops: true
crop_jpeg_quality: 90
dedup:
cooldown_seconds: 3.0
grid_size: 100.0
high_confidence_threshold: 0.75
significant_count_threshold: 5
max_tracked_per_camera: 50
# -------------------------------------------------------------------------
# Helmet Detection
# -------------------------------------------------------------------------
# Detects whether people are wearing safety helmets (construction sites, etc.).
# Triggers alerts for "no helmet" detections.
# -------------------------------------------------------------------------
helmet_detection:
enabled: false
confidence: 0.75
alert:
consecutive_threshold: 5
max_miss: 2
cooldown_secs: 10.0
tracking:
enabled: true
high_conf_threshold: 0.5
low_conf_threshold: 0.1
confirm_frames: 3
max_lost_frames: 30
data_lake:
enabled: true
bucket_prefix: helmet-detection-events
min_count: 1
jpeg_quality: 85
queue_size: 512
workers: 2
batch_size: 16
parquet_enabled: true
parquet_batch_size: 1000
save_json: false
db_enabled: true
db_batch_size: 16
db_batch_timeout_ms: 100
db_channel_size: 256
save_crops: true
crop_jpeg_quality: 90
dedup:
cooldown_seconds: 3.0
grid_size: 100.0
high_confidence_threshold: 0.75
significant_count_threshold: 5
max_tracked_per_camera: 50
# -------------------------------------------------------------------------
# Weapon Detection
# -------------------------------------------------------------------------
# Detects weapons (guns, knives, etc.) in the frame.
# Lower confidence threshold and shorter cooldown for safety-critical alerts.
# -------------------------------------------------------------------------
weapon_detection:
enabled: false
confidence: 0.5 # Lower threshold for safety-critical detection
nms_threshold: 0.45
alert:
consecutive_threshold: 5
max_miss: 2
cooldown_secs: 5.0 # Shorter cooldown for urgent alerts
tracking:
enabled: true
high_conf_threshold: 0.5
low_conf_threshold: 0.1
confirm_frames: 3
max_lost_frames: 30
data_lake:
enabled: true
bucket_prefix: weapon-detection-events
min_count: 1
jpeg_quality: 85
queue_size: 512
workers: 2
batch_size: 16
parquet_enabled: true
parquet_batch_size: 1000
save_json: false
db_enabled: true
db_batch_size: 16
db_batch_timeout_ms: 100
db_channel_size: 256
save_crops: true
crop_jpeg_quality: 90
dedup:
cooldown_seconds: 5.0 # Longer dedup for weapon events
grid_size: 100.0
high_confidence_threshold: 0.75
significant_count_threshold: 5
max_tracked_per_camera: 50
# -------------------------------------------------------------------------
# Phone Usage Detection
# -------------------------------------------------------------------------
# Two-stage: detect humans -> crop -> classify phone usage.
# Higher confidence threshold to reduce false positives on small objects.
# -------------------------------------------------------------------------
phone_usage_detection:
enabled: false
confidence: 0.8 # Higher threshold (small object, prone to FP)
nms_threshold: 0.45
crop_padding: 0.1 # Extra padding around human crop for context
alert:
consecutive_threshold: 5
max_miss: 2
cooldown_secs: 5.0
tracking:
enabled: true
high_conf_threshold: 0.5
low_conf_threshold: 0.1
confirm_frames: 3
max_lost_frames: 30
data_lake:
enabled: true
bucket_prefix: phone-usage-detection-events
min_count: 1
jpeg_quality: 85
queue_size: 512
workers: 2
batch_size: 16
parquet_enabled: true
parquet_batch_size: 1000
save_json: false
db_enabled: true
db_batch_size: 16
db_batch_timeout_ms: 100
db_channel_size: 256
save_crops: true
crop_jpeg_quality: 90
dedup:
cooldown_seconds: 30.0 # Longer cooldown to avoid spam
grid_size: 100.0
high_confidence_threshold: 0.75
significant_count_threshold: 5
max_tracked_per_camera: 50
# -------------------------------------------------------------------------
# Fight Detection
# -------------------------------------------------------------------------
# Detects physical fights/altercations between people.
# Uses pose or interaction analysis on tracked humans.
# -------------------------------------------------------------------------
fight_detection:
enabled: false
confidence: 0.5
nms_threshold: 0.45
alert:
consecutive_threshold: 5
max_miss: 2
cooldown_secs: 5.0
tracking:
enabled: true
high_conf_threshold: 0.5
low_conf_threshold: 0.1
confirm_frames: 3
max_lost_frames: 30
data_lake:
enabled: true
bucket_prefix: fight-detection-events
min_count: 1
jpeg_quality: 85
queue_size: 512
workers: 2
batch_size: 16
parquet_enabled: true
parquet_batch_size: 1000
save_json: false
db_enabled: true
db_batch_size: 16
db_batch_timeout_ms: 100
db_channel_size: 256
save_crops: true
crop_jpeg_quality: 90
dedup:
cooldown_seconds: 3.0
grid_size: 100.0
high_confidence_threshold: 0.75
significant_count_threshold: 5
max_tracked_per_camera: 50
# -------------------------------------------------------------------------
# People Counting (migrated from builtin to dynamic plugin)
# -------------------------------------------------------------------------
people_counting:
enabled: false
confidence: 0.25
nms_threshold: 0.45
tracking:
enabled: true
high_conf_threshold: 0.5
low_conf_threshold: 0.1
confirm_frames: 3
max_lost_frames: 30
data_lake:
enabled: true
bucket_prefix: people-counting-events
min_count: 1
jpeg_quality: 85
queue_size: 512
workers: 2
batch_size: 16
parquet_enabled: true
parquet_batch_size: 1000
save_json: false
db_enabled: true
db_batch_size: 16
db_batch_timeout_ms: 100
db_channel_size: 256
save_crops: true
crop_jpeg_quality: 90
dedup:
cooldown_seconds: 3.0
grid_size: 100.0
high_confidence_threshold: 0.75
significant_count_threshold: 5
max_tracked_per_camera: 50
# -------------------------------------------------------------------------
# License Plate Recognition (migrated from builtin to dynamic plugin)
# -------------------------------------------------------------------------
license_plate_recognition:
enabled: false
confidence: 0.5
nms_threshold: 0.45
tracking:
enabled: true
high_conf_threshold: 0.5
low_conf_threshold: 0.1
confirm_frames: 2
max_lost_frames: 15
data_lake:
enabled: true
bucket_prefix: license-plate-recognition-events
min_count: 1
jpeg_quality: 85
queue_size: 512
workers: 2
batch_size: 16
parquet_enabled: true
parquet_batch_size: 1000
save_json: false
db_enabled: true
db_batch_size: 16
db_batch_timeout_ms: 100
db_channel_size: 256
save_crops: true
crop_jpeg_quality: 90
dedup:
cooldown_seconds: 10.0
grid_size: 100.0
high_confidence_threshold: 0.75
significant_count_threshold: 5
max_tracked_per_camera: 50
# -------------------------------------------------------------------------
# Traffic Violation Detection (migrated from builtin to dynamic plugin)
# -------------------------------------------------------------------------
traffic_violation_detection:
enabled: false
confidence: 0.5
nms_threshold: 0.45
tracking:
enabled: true
high_conf_threshold: 0.5
low_conf_threshold: 0.1
confirm_frames: 3
max_lost_frames: 30
data_lake:
enabled: true
bucket_prefix: traffic-violation-events
min_count: 1
jpeg_quality: 85
queue_size: 512
workers: 2
batch_size: 16
parquet_enabled: true
parquet_batch_size: 1000
save_json: false
db_enabled: true
db_batch_size: 16
db_batch_timeout_ms: 100
db_channel_size: 256
save_crops: true
crop_jpeg_quality: 90
dedup:
cooldown_seconds: 3.0
grid_size: 100.0
high_confidence_threshold: 0.75
significant_count_threshold: 5
max_tracked_per_camera: 50
# =============================================================================
# Logging
# =============================================================================
logging:
enabled: true
dir: logs # Log files directory (relative to executable)
max_days: 7 # Auto-cleanup logs older than N days
|