ALPR Parking Counting

The ALPR Parking Time service is designed for mobile camera systems mounted inside moving vehicles. It enables parking zone monitoring, parking time enforcement, and vehicle presence tracking by continuously scanning and recognizing license plates while driving through designated areas.

Use Cases

• Parking zone enforcement - Monitor vehicle parking duration in paid zones

• Parking lot management - Track vehicle entry/exit and dwell times

• Street parking audits - Survey parked vehicles along routes

• Permit verification - Identify vehicles with/without valid parking permits

• Time-limited parking monitoring - Detect overstays in restricted areas

Camera Requirements

In-Vehicle Mounted Cameras

Cameras installed behind the windshield or on vehicle roof, scanning parked vehicles while driving.

Video requirements:

• Frame rate: 50-60 FPS (higher frame rate is essential)

• Resolution: Full HD (1920×1080) minimum

• Vehicle speed: < 30 km/h for reliable recognition

• Wide-angle lens recommended for side-parked vehicles

Why high frame rate?

• Moving camera + stationary plates = motion blur risk

• Higher FPS provides more frames per vehicle

• Increases chance of capturing sharp, readable plate image

Mounting considerations:

• Stable mounting to minimize vibration

• Anti-glare filters for windshield mounting

• Multiple cameras for left/right side coverage

Generated Reports and Data

The parking time service provides specialized reports for parking analysis.

Output Files

File	Description
History.csv	Complete scan history with timestamps
Alpr.csv	All license plate readings with metadata
AlprOccurences.csv	Plate occurrence matrix by time intervals
AlprParktime.csv	Parking duration analysis per vehicle
Alpr0to500.xlsx, ...	Excel sheets with images

Parking Time Analysis

The AlprParktime.csv file contains calculated parking durations:

Field	Description
License Plate	Recognized plate text
Start	First detection timestamp
End	Last detection timestamp
Duration	Calculated parking time (minutes)

This data is grouped by default 5-minute intervals and sorted by duration (longest first).

Data Fields

Each license plate detection includes:

Field	Description
ALPR	Recognized license plate text
ALPR_score	Recognition confidence (0.0 - 1.0)
ALPR_type	License plate type classification
ALPR_region	Region/country code
ALPR_lettercode	District/city letter code
Czas (Time)	Detection timestamp
Image	Vehicle image path
ALPR_image	Cropped plate image path

Occurrence Matrix

The AlprOccurences.csv provides a time-based matrix showing when each license plate was detected:

Time	ABC1234	XYZ5678	...
08:00	1	0	...
08:05	1	1	...
08:10	0	1	...

This matrix helps visualize vehicle presence patterns across time.

Workflow Example

1. Route planning - Define patrol route through parking zone

2. Data collection - Drive route with camera recording (repeat every X minutes)

3. Processing - Upload videos to AITracker for ALPR analysis

4. Parking time calculation - System correlates same plates across multiple passes

5. Enforcement report - Export vehicles exceeding time limits

Accuracy

Factor	Impact on Accuracy
Vehicle speed < 30 km/h	Optimal
Vehicle speed 30-50 km/h	Reduced accuracy
Night-time operation	Requires IR camera
Angle to parked vehicles	15-45° optimal

Overall ALPR accuracy: ~90% under optimal conditions

Processing speed: x1-x2 real-time for 50-60 FPS video

Configuration

Configure the parking time service:

[Service]
id = "ServiceAlprParktime"

[ALPR]
enabled = true
save_images = true
grouping_period = "5min"  # Time grouping for occurrence analysis

Limitations

• GPS integration - Currently timestamps only; GPS correlation requires external tools

• Multiple passes - Parking time calculation requires multiple route passes

• Plate visibility - Obstructed or dirty plates may not be recognized