File size: 13,677 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 | [plugin]
name = "traffic_violation_detection"
version = "5.3.0"
display_name = "Phát hiện vi phạm giao thông"
description = "Phát hiện vi phạm giao thông: vượt đèn đỏ, sai làn, ngược chiều, không đội mũ, quay đầu xe, quá tốc độ"
library = "traffic_violation_detection.dll"
[capabilities]
use_gpu = true
use_batching = true
use_tracking = true
use_zone = true
[classes]
names = ["car", "motorbike", "truck", "bus"]
defaults = ["car", "motorbike", "truck", "bus"]
[models]
# Order matters: GUI's "Models" section binds model_1 → files[0], etc.
files = [
"mbfs_vehicle_det_v4.onnx", # model_1 — Vehicle Detector
"traffic-light-det.onnx", # model_2 — Traffic Light Detector
"helmet_dt_v10.onnx", # model_3 — Helmet Detector
"lp_detection_v4.onnx", # model_4 — Plate Detector
"license_place_rec_crnn_v8.onnx", # model_5 — CRNN/CTC sequence OCR (v8)
"weight_detect_corner.onnx", # model_6 — 4-corner plate detector
"mbfsn_cls_brand_car_v1.onnx", # model_7 — Joint Make_Model classifier (80 classes, single softmax)
]
[inference]
model = "mbfs_vehicle_det_v4.onnx"
confidence = 0.5
nms = 0.45
input_size = 640
[models.traffic_light]
model = "traffic-light-det.onnx"
kind = "yolo"
input_size = 640
confidence = 0.5
nms = 0.45
[models.helmet]
model = "helmet_dt_v10.onnx"
kind = "yolo"
input_size = 640
confidence = 0.5
nms = 0.45
[models.plate_detector]
model = "lp_detection_v4.onnx"
kind = "yolo"
input_size = 640
confidence = 0.3
nms = 0.45
# CRNN/CTC sequence OCR (v8). Single model for both 1-line and 2-line plates.
# See LPR manifest for full algorithm description (corner-warp + LPTR-AFLNet).
[models.plate_ocr_crnn]
model = "license_place_rec_crnn_v8.onnx"
kind = "ocr_seq"
input_size = [128, 32] # [width, height]
[models.plate_ocr_crnn.ocr_seq]
vocab = "0123456789ABCDEFGHKLMNPRSTUVXYZ"
channels = 1 # grayscale; host applies BT.601 luma + x/255.0
# 4-corner plate corner detector (perspective rectification source).
[models.plate_corner]
model = "weight_detect_corner.onnx"
kind = "yolo"
input_size = 320
confidence = 0.3
nms = 0.5
# Joint Make_Model classifier (YOLOv8n-cls, 80 classes "{Make}_{Model}").
# Single softmax head; plugin splits label on the first '_' to recover
# brand vs model. Replaces the older two-head model (brand_logits 9 +
# model_logits 72) — same datalake schema preserved.
[models.vehicle_brand]
model = "mbfsn_cls_brand_car_v1.onnx"
kind = "classifier"
input_size = 224
# TEMP — pixel-based traffic-light classifier.
#
# Substitutes for the YOLO `traffic_light` model (which is a single-class
# "Red" detector and hallucinates red on green phases at cam 301). When a
# multi-class traffic-light model is retrained, delete this section and
# revert the `light_pixel` swap in `lib.rs` Phase B1.
#
# All defaults tuned against 227 LPR frames spanning cams 1/4/5; operators
# rarely need to touch them.
[traffic_light_pixel]
value_threshold = 130 # 0–255; min max(R,G,B) for a pixel to count as "lit"
red_margin = 60 # R - max(G,B) for a pixel to count as red
green_margin = 40 # G - max(R,B) for a pixel to count as green
yellow_margin = 50 # min(R,G) - B for a pixel to count as yellow
min_lit_fraction = 0.02 # winning color must be at least this fraction of ROI
# ── Config page schema (data-driven GUI) ──────────────────────────────────
[[config.sections]]
title = "MODELS"
fields = [
{ key = "model_1", type = "model", label = "Vehicle Detector:" },
{ key = "model_2", type = "model", label = "Light Classifier:" },
{ key = "model_3", type = "model", label = "Helmet Classifier:" },
{ key = "model_4", type = "model", label = "Plate Detector:" },
{ key = "model_5", type = "model", label = "CRNN OCR Model (v8):" },
{ key = "model_6", type = "model", label = "Corner Detector:" },
{ key = "model_7", type = "model", label = "Vehicle Brand:" },
]
[[config.sections]]
title = "DETECTION"
fields = [
{ key = "inference.confidence", type = "slider", label = "Vehicle Detection Confidence", help = "Primary YOLO vehicle detection confidence (0.0-1.0)", min = 0.0, max = 1.0, step = 0.05, default = 0.5 },
{ key = "models.traffic_light.confidence", type = "slider", label = "Traffic Light Confidence", help = "Traffic light detection confidence (0.0-1.0)", min = 0.0, max = 1.0, step = 0.05, default = 0.5 },
{ key = "models.helmet.confidence", type = "slider", label = "Helmet Detection Confidence", help = "Helmet detection confidence (0.0-1.0)", min = 0.0, max = 1.0, step = 0.05, default = 0.5 },
{ key = "models.plate_detector.confidence", type = "slider", label = "Plate Detection Confidence", help = "Plate detector confidence (0.0-1.0)", min = 0.0, max = 1.0, step = 0.05, default = 0.3 },
{ key = "two_line_aspect_max", type = "slider", label = "2-Line Plate Aspect Threshold", help = "Plates with width/height below this are tried as 2-line first. Falls back to 1-line on failed validation.", min = 1.5, max = 3.5, step = 0.1, default = 2.3 },
{ key = "lpr_min_ocr_confidence", type = "slider", label = "Min Plate OCR Confidence", help = "Reject whole plate readings below this averaged per-character confidence (0.0-1.0)", min = 0.0, max = 1.0, step = 0.05, default = 0.3 },
]
[[config.sections]]
title = "VEHICLE IDENTIFICATION"
fields = [
{ key = "vehicle_brand_enabled", type = "toggle", label = "Brand/Model Classifier", description = "Classify make and model on confirmed vehicle tracks", default = true },
{ key = "vehicle_brand_classify_interval", type = "number", label = "Classify Interval", help = "Frames between brand+model classification (1-30)", min = 1, max = 30, step = 1, default = 5, visible_when = "vehicle_brand_enabled" },
{ key = "vehicle_brand_min_confidence", type = "slider", label = "Min Brand Confidence", help = "Drop brand+model predictions below this joint-softmax confidence (80-class Make_Model head). 0.3 is a reasonable default; both brand and model receive the same confidence.", min = 0.0, max = 1.0, step = 0.05, default = 0.3, visible_when = "vehicle_brand_enabled" },
{ key = "vehicle_model_min_confidence", type = "slider", label = "Min Model Confidence", help = "Legacy knob from the two-head classifier era. The new joint Make_Model classifier uses vehicle_brand_min_confidence only; this slider is currently inert.", min = 0.0, max = 1.0, step = 0.05, default = 0.1, visible_when = "vehicle_brand_enabled" },
]
[[config.sections]]
title = "TRACKING (ByteTrack)"
fields = [
{ key = "tracking.high_conf_threshold", type = "float", label = "High Confidence", help = "High confidence threshold (0.1-1.0)", min = 0.1, max = 1.0, step = 0.05, default = 0.5 },
{ key = "tracking.low_conf_threshold", type = "float", label = "Low Confidence", help = "Low confidence threshold (0.01-0.5)", min = 0.01, max = 0.5, step = 0.01, default = 0.1 },
{ key = "tracking.confirm_frames", type = "number", label = "Confirm Frames", help = "Frames to confirm track (1-10)", min = 1, max = 10, step = 1, default = 3 },
{ key = "tracking.max_lost_frames", type = "number", label = "Max Lost Frames", help = "Frames before track is lost (1-120)", min = 1, max = 120, step = 1, default = 30 },
]
[[config.sections]]
title = "SCHEDULING"
fields = [
{ key = "light_classify_interval", type = "number", label = "Light Classify Interval", help = "Frames between traffic light classification (1-10)", min = 1, max = 10, step = 1, default = 2 },
{ key = "general_classify_interval", type = "number", label = "General Classify Interval", help = "Frames between general classification (1-10)", min = 1, max = 10, step = 1, default = 3 },
{ key = "max_trajectory_points", type = "number", label = "Max Trajectory Points", help = "Maximum trajectory history per track (10-200)", min = 10, max = 200, step = 10, default = 60 },
]
[[config.sections]]
title = "VIOLATION RULES"
# Toggles here promote to per-camera overrides: operators can enable
# red-light detection on cam 16 but not cam 17, etc. Everything else in
# the manifest stays plugin-wide (default `per_camera = false`).
per_camera = true
fields = [
# ── Red Light Running ──
{ key = "rules.red_light.enabled", type = "toggle", label = "Red Light Running", description = "Detect vehicles crossing stop line during red signal", default = true },
{ key = "rules.red_light.yellow_grace_period_ms", type = "number", label = "Yellow Grace (ms)", help = "Grace period after yellow light (0-2000)", min = 0, max = 2000, step = 100, default = 500, visible_when = "rules.red_light.enabled" },
{ key = "rules.red_light.confirm_window", type = "number", label = "Confirm Window", help = "Sliding window size (1-30)", min = 1, max = 30, step = 1, default = 5, visible_when = "rules.red_light.enabled" },
{ key = "rules.red_light.confirm_threshold", type = "number", label = "Confirm Threshold", help = "Required confirmations (1-30)", min = 1, max = 30, step = 1, default = 3, visible_when = "rules.red_light.enabled" },
# ── Wrong Lane ──
{ key = "rules.wrong_lane.enabled", type = "toggle", label = "Wrong Lane", description = "Detect vehicles driving in restricted lanes", default = true },
{ key = "rules.wrong_lane.confirm_window", type = "number", label = "Confirm Window", help = "Sliding window size (1-30)", min = 1, max = 30, step = 1, default = 5, visible_when = "rules.wrong_lane.enabled" },
{ key = "rules.wrong_lane.confirm_threshold", type = "number", label = "Confirm Threshold", help = "Required confirmations (1-30)", min = 1, max = 30, step = 1, default = 3, visible_when = "rules.wrong_lane.enabled" },
# ── Wrong Direction ──
{ key = "rules.wrong_direction.enabled", type = "toggle", label = "Wrong Direction", description = "Detect vehicles traveling against permitted direction", default = true },
{ key = "rules.wrong_direction.smoothing_window", type = "number", label = "Smoothing Window", help = "Direction smoothing window (1-30)", min = 1, max = 30, step = 1, default = 5, visible_when = "rules.wrong_direction.enabled" },
{ key = "rules.wrong_direction.confirm_window", type = "number", label = "Confirm Window", help = "Sliding window size (1-30)", min = 1, max = 30, step = 1, default = 7, visible_when = "rules.wrong_direction.enabled" },
{ key = "rules.wrong_direction.confirm_threshold", type = "number", label = "Confirm Threshold", help = "Required confirmations (1-30)", min = 1, max = 30, step = 1, default = 4, visible_when = "rules.wrong_direction.enabled" },
# ── No Helmet ──
{ key = "rules.no_helmet.enabled", type = "toggle", label = "No Helmet", description = "Detect motorcycle riders without helmets", default = true },
{ key = "rules.no_helmet.confirm_window", type = "number", label = "Confirm Window", help = "Sliding window size (1-30)", min = 1, max = 30, step = 1, default = 5, visible_when = "rules.no_helmet.enabled" },
{ key = "rules.no_helmet.confirm_threshold", type = "number", label = "Confirm Threshold", help = "Required confirmations (1-30)", min = 1, max = 30, step = 1, default = 3, visible_when = "rules.no_helmet.enabled" },
# ── Illegal U-Turn ──
{ key = "rules.illegal_uturn.enabled", type = "toggle", label = "Illegal U-Turn", description = "Detect unauthorized U-turns in restricted zones", default = true },
{ key = "rules.illegal_uturn.turn_threshold_degrees", type = "float", label = "Turn Threshold (°)", help = "Minimum heading change for U-turn (90-180)", min = 90.0, max = 180.0, step = 5.0, default = 150.0, visible_when = "rules.illegal_uturn.enabled" },
{ key = "rules.illegal_uturn.trajectory_window", type = "number", label = "Trajectory Window", help = "Frames for trajectory analysis (5-120)", min = 5, max = 120, step = 5, default = 30, visible_when = "rules.illegal_uturn.enabled" },
{ key = "rules.illegal_uturn.confirm_window", type = "number", label = "Confirm Window", help = "Sliding window size (1-30)", min = 1, max = 30, step = 1, default = 5, visible_when = "rules.illegal_uturn.enabled" },
{ key = "rules.illegal_uturn.confirm_threshold", type = "number", label = "Confirm Threshold", help = "Required confirmations (1-30)", min = 1, max = 30, step = 1, default = 3, visible_when = "rules.illegal_uturn.enabled" },
# ── Speeding ──
{ key = "rules.speeding.enabled", type = "toggle", label = "Speeding", description = "Detect vehicles exceeding speed limit (requires calibration)", default = false },
{ key = "rules.speeding.speed_limit_kmh", type = "float", label = "Speed Limit (km/h)", help = "Maximum allowed speed (10-200)", min = 10.0, max = 200.0, step = 5.0, default = 60.0, visible_when = "rules.speeding.enabled" },
{ key = "rules.speeding.measurement_window", type = "number", label = "Measurement Window", help = "Frames for speed measurement (3-60)", min = 3, max = 60, step = 1, default = 10, visible_when = "rules.speeding.enabled" },
{ key = "rules.speeding.confirm_window", type = "number", label = "Confirm Window", help = "Sliding window size (1-30)", min = 1, max = 30, step = 1, visible_when = "rules.speeding.enabled" },
{ key = "rules.speeding.confirm_threshold", type = "number", label = "Confirm Threshold", help = "Required confirmations (1-30)", min = 1, max = 30, step = 1, visible_when = "rules.speeding.enabled" },
]
|