Automatic number plate recognition systems
In this world, traffic control and monitoring systems became a common and central issue for the majority of developing countries. The number of vehicles is rapidly increasing, so the number of violation are increasing day by day. Quickly stealing cars and other wide ranges of the spectrum of violati
2025-06-28 16:25:26 - Adil Khan
Automatic number plate recognition systems
Project Area of Specialization Information & Communication TechnologyProject Summary Project summary:In this world, traffic control and monitoring systems became a common and central issue for the majority of developing countries. The number of vehicles is rapidly increasing, so the number of violation are increasing day by day. Quickly stealing cars and other wide ranges of the spectrum of violations on the road have now become common things in our daily life.
Background and IntroductionAutomatic number plate recognition systems are already implemented in many other countries like India, Australia, Korea, and a few others. With the help of the early development of ANPR systems, strict implementation of license plate standards in these countries plays a very important role. Such type of system uses specific standard features of the plate such as Dimension of the plate, border for the plate, color and font of the characters, etc. and such type of issues are also faced by some common places like automatic toll tax collection, car park system character segmentation and character recognition.
After that the automatic number plate recognition system is efficient framework, but we have still some issues that we are facing in our system will discuss one by one. Pakistan faces some common issues with the development of such a system in Pakistan. Pakistan have not follow any fixed template and design of number plates and there is no fixed position of the number plate on the vehicles, under different lighting conditions detection and recognition of text from number plates are also an issue in Pakistan. These all are the challenges tasks for us in our system. We are going to design our system in such a way that we can overcome these issues.
After working on the different framework, we will make a framework which gives us better result in low computer processing speed and time. Our system based on six phases. Which has the following steps.
Number plate image capture (by using any camera).
Number plate extraction from the image.
Image preprocessing (specially designed for Pakistan Vehicles Number plate because Pakistan has no fixed template for their vehicles number plate)
Character recognition
Number plate management process
Train our model to give a matching result of the number plate from our .xml file. Data set.
Project Objectives Aims and objectives:Our aim is to develop an Automatic number plate recognition system.
There are too many traffic-related problems are present. Especially control and monitoring systems became a common and central issue for many countries. After that, some other issues are very common in Pakistan. Pakistan does have not a proper number plate template and proper location of the number plate on the vehicles which would make the detection of vehicle plates a difficult task for us. We would develop such a framework we can overcome these issues.
- Detection of the number plates of different designs and templates.
- Detect the number plate which is located anywhere on the cars.
- Recognize the text from the number plate under different lighting conditions, under any quality of the image.
In the ANPR system, we capture the image of the vehicle that looks easy but it is a quite difficult process to take the image of moving vehicles in real-time. In such a manner that none of the components of the vehicle especially the number plate of the vehicles wouldn't be missed. Now Automatic number plate recognition system is a key aspect to resolve such a time problem. However, to create such a time system, we will be needed to develop such time algorithms which will get the image by using any camera and then identify the location of the number plate region from the image, extract the characters from it, and recognize the character by using different built-in libraries that written in python.
Project Implementation Method Modeling and description:Our system based on six phases. Which has the following steps.
- Number plate image capture (by using any camera).
- Number plate extraction from the image.
- Image preprocessing (specially designed for Pakistan Vehicles Number plate because Pakistan has no fixed template for their vehicles number plate)
- Character recognition
- Number plate management process
- Train our model to give a matching result of the number plate from our .xml file. Data set.
| Importing libraries |
| Defining the path of pytesserect library |
| Declare the path of pak.xml |
| Function for image preprocessing |
| Cv2.imread |
| Detect the Multi-scale |
| Image to Grayscale |
| Remove noise/blur |
| Resize image |
| Declare the region of interest (plate) |
| Adaptive threshold |
| Use kernel (Display) |
| Pass from (dilate/erode) |
| Plate to Grayscale |
| Character recognition |
| Cropped plate |
| Display text (AVE068) |
| Pytesserect |
| Image to text |
| Plate to Grayscale |
| Saving Data |
| Use file open function |
| Display text (AVE068) |
| Use f. write (text) |
| Written in .txt file |
| Open .txt file in write mode |
Importing libraries
Defining the path of pytesserect library
Declare the path of pak.xml
Function for image preprocessing
Cv2.imread
Detect the Multi-scale
Image to Grayscale
Remove noise/blur
Resize image
Declare the region of interest (plate)
Adaptive threshold
Use kernel
(Display)
Pass from
(dilate/erode)
Plate to Grayscale
Character recognition
Cropped plate
Display text
(AVE068)
Pytesserect
Image to text
Plate to Grayscale
Saving Data
Use file open function
Display text
(AVE068)
Use f. write (text)
Written in .txt file
Open .txt file in write mode
Benefits of the Project Benefits of the Project:We are developing automatic number plate recognition system which has so many benefits that are given below.
Automatic number plate recognition system will change the era of monitoring of cars in different places such has parking lots, toll plazas motorways etc. This help us to easy track the vehicles and recognize them through their license number plate and easy keep the record in our database with less amount of error because manually monitoring and data entry was a challenge task .this system also reduce the crime rate in which cars are included.
In Pakistan we develop ANPR system which is working on different design of number plate (which is the challenge task for Pakistan) but we are trying to overcome this issue. Which will be helpful for different department in Pakistan.
Technical Details of Final Deliverable Technical Details of Final Deliverable:our system working on different condition such as detect the number plate from image and video as well first of all we specify the path of image and video in code after passing through different image and video preprocessing steps it will generate an output of cropped number plate in the form of image. with the help of python library we can show our final result on the screen in the form of three images, first image show the grayscale image, second image how the cropped number plate image and third image shows the image of whole vehicle along with bounding boxes around the number plate along with text that can be recognized by the library which we are using in code. We can also get the same result for the video part. The output that was generated by the code are written in the excel file and text file with the proper sequences and will stored the image of the cropped number plate in our system as well. This is how our system worked on user end without any user interface.
| Defining the path of pytesserect library |