User Guide

1. Creating a camera configuration file and setting up the GUI

Starting the program and setting up the camera configuration file is a fairly straight forward process. But, the format of the configuration file varies according to where we run the program

Ubuntu GUI

How to set up the camera configuration file

1. Create a text file ( With any name)

2. Content of the config file should consist of one camera port or video file path per row as shown below

0 
http://192.168.0.1:8080 
rstp://iapp:i@pp2021@192.168.1.49/MediaInput/h264
path/video_name.mp4

Starting the program:

python3 camera_pipline_GUI.py 

This will take you to the GUI shown below

Ubuntu Terminal

How to set up the camera configuration file

1. Create a text file (Recommended name: camera_config_server.txt)

2. Content of the config file should look like below

Example of a camera configuration file

[port or path] [name] [vehicle_size] [time to save the results] [detect color]

Detect color:

False: will not detect color

True: will detect color (process will be slower)

Starting the program:

With recommended name (camera_config_server.txt):

python3 camera_pipline_serverRun.py

With any other custom name:

python3 camera_pipline_serverRun.py --source [path to the configuration file]

Dummy Camera-configuration file

0 front_camera 2.5 3 False
http://192.168.0.1:8080 back_camera 2.8 2 True
rstp://iapp:i@pp2021@192.168.1.49/MediaInput/h264 my_camera 2 1 False
path/video_name.mp4 my_video 2.5 1 False
path/gas_station.mp4 gas_station 2.8 1 True

Do not leave spaces in-between camera_name

2. How to choose the number of seconds for results to save

1 second = Results of 30 successful licence plate OCR's completed

So, every 1 second results of 30 vehicles will be summarized and saved in the database. The number of unique vehicle appear in 30 frames is more likely to be minimum of one and maximum of three.

Recommended

4-10 seconds for cameras (live streaming) - summary of 120-300 vehicle frames will be saved to the database.

1-5 seconds for videos of 10-20 minutes long - summary of 30-150 vehicle frames will be saved to the database.

3. How to choose the vehicle size ratio

The size ratio of the vehicle you choose on the GUI (if run on Ubuntu GUI) or camera config file (if run on Ubuntu terminal) will be compared with the vehicle size ratio of each vehicle on the frame/screen as shown below.

REQUIREMENT :

the_actual_vehicle_ratio <= (less than or equal) the_vehicle_ratio_you_choose

Note: The system will only allow the vehicles which passes the above requirement to the next step of the LPR system.

Check out the examples below for further understanding on how to choose the vehicle ratio,

Example - 1

From the above example according to the vehicle size ratio given by the user only one vehicle (Pickup_Truck - ISUZU - Black) will be proceed to the next step of the LPR system.

Example - 2

From the above example according to the vehicle size ratio given by the user only one vehicle (Car - Mercedes-Benz - Gray) will be proceed to the next step of the LPR system.