[Python ANPR Course] Real-Time Automatic Number Plate Recognition In Your Car!

A tutorial on how to code a real-time Automatic Number Plate Recognition (ANPR) system that you can install in your vehicle! You'll basically be a highway patrol police car!
This is a 56 minute Python course on how to develop an Automatic License Plate Recognition (ALPR) system using the Asus Tinker Edge T with Python 3.
Source code here: github.com/edgecate/ANAL
You'll learn how to:
✅ Use a Deep Learning AI model to detect license plates,
✅ Photograph it with OpenCV 4,
✅ Convert the photo to text with OpenALPR,
✅ Display the results on to a GUI with PyQT5 & Pandas, and
✅ Install the system in your car
Chapters:
0:00 Introduction
0:53 Pre-requisites
1:52 Create The GUI
4:58 Export The GUI
6:01 Add Webcam Video To GUI
12:16 Setup GUI Table
14:35 Copy .PY To Tinker Edge T
19:54 Add Object Detection
30:06 Save License Plate As An Image
39:27 Convert License Plate Image To Text
45:03 Check Registration
47:25 Output Results To CSV
53:31 Car Installation
55:26 Outro
For Hardware, you'll need:
-an Asus Tinker Edge T with a power supply & HDMI cable,
-a webcam,
-a portable monitor, and a
-3-5 metre USB cable
For Software, the libraries used for this course are:
-OpenCV 4,
-OpenALPR,
-PyQT5,
-TensorFlow Lite,
-Tesseract OCR,
-ImageMagick,
-Jupyter,
-Chromium, and
-Pandas.
Credits & Links:
PyQt5 OpenCV WebCam Using QThread: stackoverflow.com/questions/6...
Install Mendel OS: • Asus Tinker Edge T - I...
Install Chromium: • Asus Tinker Edge T - I...
Install Jupyter: • Asus Tinker Edge T - I...
Install OpenCV 4: • Asus Tinker Edge T - B...
Install Tesseract OCR: • Asus Tinker Edge T - B...
Install OpenALPR: • Asus Tinker Edge T - B...

[@blacksabbathbr]

I've heard ANPR, ALPR and LPR but never ANAL. Thanks for the giggle

[@robotdog8713]

Thank alot,Your code is valid. I'm happy to support.

[@mohsenimani6652]

Great video, quick question, how did you get the model? did you train it or it's a pretrained model? thanks

[@redsandro]

I was trying a dash cam approach with the native Google Coral 5MP Camera module, but the Asus Tinker Edge T really can't store/encode video. I believe they chose their GPU poorly as the Vivante GC7000 Lite doesn't have h264 hardware encoding. The 2012 Raspberry Pi has no problem processing its camera module feed into h264 for dash cam storage. The Broadcom VideoCore IV GPU has dedicated hardware h264 encoding. Even the $5 Raspberry Pi Zero has this GPU. Am I missing something? Or did Asus really (really) choose poorly?

[@priyamdalwadi8865]

Amazing video!!. Is there any other way to deal with duplicates created of crop images at 30Fps camera?

[@SahlEbrahim]

can you implement this on the phone?? without the extra hardware??

[@adityagaikwad2856]

Would this setup work to detect the number plate of any pattern? As I am from India, there's a different pattern of number plates here. would it recognize that?

[@khushhalgupta3997]

sir while installing pyqt5 tools its showing error

[@djamaltoe3774]

is it obliged to use edgetpu ?? can't we use our computer only without an other hardware ??

[@Dramon11]

Where can I find documentation for the asus edge tiker R ???

[@AdrianFrasineanu]

Hello! Is it possible to run this software on a web platform?

[@ivargasmx]

Hello, first of all, thank you for all the information and dedication you share on your excellent channel. I want to ask you what you recommend to make an application for the detection and recognition of Car License Plates/License Plates/Patents (ANPR) of a WEB camera or a standard video surveillance camera for a development board of medium capacity but in real time such as to give access to cars at a door or barrier (night lighting may be adequate). With the following options and know which one you recommend (if you have any other, welcome the suggestion) - Development board: Raspberry Pi, Google Coral, ASUS Thinker Edge, Jetson Nano, etc. - Programming Language: Python, C++, Java, .Net, Nod.js, etc. - Libraries and/or Framework / tools: Yolo, OpenCV, OCR, preferably not Cloud - To train the model that you would recommend, or if there is already one that can be purchased for my case - I am in Mexico and the license plates would only be from the country

[@gametime622]

Will it work amd gpu and windows 10 on my pc. Please let me know.

[@djamaltoe3774]

No module named 'pycoral.adapters' please help me

[@abishek7697]

Could you send the requirements.txt file So that all the coders will be able to have a similar environment to yours

[@anwarabouhamze2381]

is it possible to use ipcam insted an analog cam

[@javiscript8246]

No module named 'pycoral.adapters'

[@robotdog8713]

Support on raspberry pi 4b or not.

[@oferorly]

Hi I am looking for system that read LP for a Non-motorized Vehicle E bike and E Scooters I will be Happy to here if you heard about something like it

[@diegonp12]

Is there any way to make this system without the asus thinker edge t card ?

[@edgecate]

Yup. All the steps here should work on any other device that runs Python 3. You might have to use another model though. I'm not sure if my TF Lite model works on non-Edge TPU devices, but it's unconfirmed. Either way, you can use other existing models or train your own (I have another video on that).

[@diegonp12]

@@edgecate Thanks, how could i use it on windows?

[@edgecate]

@@diegonp12 good question - that's probably a video in itself, but in short, you'd code and run it on your windows machine. You'd have to have python 3 + the python libraries installed that are mentioned in the video. Hope that helps!

[@diegonp12]

@@edgecate in the case, would I have to change just the tflite to use in windows? for testing with the webcam! Thanks for answering

[@edgecate]

​@@diegonp12 yep you'll have to do that too. You can use another model from here: github.com/tensorflow/models/blob/master/research/object_detection/g3doc/tf1_detection_zoo.md (There's also a TF2 folder, too) On top of that, just having a quick look at my code now, if you want to use it on Windows, you'll need to replace any lines of my code that use pycoral libraries (i.e., input_size, get_objects, read_label_file, make_interpreter, run_inference) with non-Edge TPU equivalent - I don't know what the non-Edge TPU equivalent would be though.

[@djamaltoe3774]

is it obliged Copy .PY To Tinker Edge T ? i juste want to do an exe Please answer me

[@edgecate]

The trained model will only work on boards with an Edge TPU. I wouldn't try to make a .exe out of this - too many dependencies.

[@djamaltoe3774]

@@edgecate Okay i see thank you very much, but i need some advices from you..

[@keske9]

0:06 window title :^)

[@yashaswigajbhiye6057]

Can you share the source code

[@edgecate]

Oops thank-you for the reminder! You can find it here: github.com/edgecate/ANAL

[@yashaswigajbhiye6057]

@@edgecate thank you

[@mauriciorodriguez1278]

Me da problemas con pycoral File "c:\Users\Cleo\Desktop\ANAL-main\anal.py", line 9, in from pycoral.adapters.common import input_size ModuleNotFoundError: No module named 'pycoral.adapters'

[@MegaAmafia]

numpy.core._multiarray_umath missing

[@chakraphatsudjai5014]

Traceback (most recent call last): File "c:/Users/Tonkla/Desktop/ANAL-main/ANAL-main/anal.py", line 12, in from pycoral.utils.edgetpu import make_interpreter File "C:\Users\Tonkla\AppData\Local\Programs\Python\Python37\lib\site-packages\pycoral\utils\edgetpu.py", line 24, in from pycoral.pybind._pywrap_coral import GetRuntimeVersion as get_runtime_version ImportError: DLL load failed: The specified module could not be found.

[@karanbhardwaj5334]

Same error