Smart toll tax collection system using Matlab for vehicle number plate recognition

Abstract:

Real time “Character Recognition & Pattern Recognition” is a latest research topic. Research Engineers all around the world are working on it. This technique is being used in many daily life applications to make human life much easy. Vehicle Number Recognition (VNR) is an image processing technology which uses efficient algorithms to detect the vehicle number from real time images. A toll road is a privately or publicly built road for which one pays a toll tax for its use. The building or facility in which a toll is collected may be called a toll plaza, toll station, or toll gate. The purpose of the research project (Smart Toll Tax Collection System Using Automatic Vehicle Number Recognition) is to automate the toll collection and to facilitate the toll collecting authorities. The system first detects the vehicle and then captures the image of the front view of the vehicle. Vehicle number plate is localized & characters are segmented. The system is designed for grayscale images so it detects the number plate regardless of its color. Template matching technique is used for the character recognition. The resulting vehicle number is then compared with the available database of all the vehicles so as to come up with information about the vehicle type & to charge toll tax accordingly. The system is then allowed to open road barrier for the vehicle & generate toll tax receipt. The vehicle information (such as passing time, date, toll amount) is also stored in the different database to maintain the record.

Algorithm for Automatic Vehicle Number Recognition is designed in MATLAB. The road barriers & traffic is controlled by using Microcontroller. System also contains LCD display to guide vehicles. Apart from toll collection system can also be used for security and authorized vehicles entrance in organizations & for automatic commercial parking systems. The proposed AVNR algorithm consists of some stages. In the first stage, Vertical edges of the acquired image are found. In the next stage, histogram analysis of its bit density is used for finding the possible candidate regions of license plate. Then in 3rd step deviation in rows is applied to locate exact candidate region. In the next stage, exact location of Number Plate is extracted by using some Morphological operators. Then in next step image region properties are found and by using bounding box Number Plate is segmented and each character is separated. In the last step each segmented image is compared with characters & digits templates which are already stored on computer & on the basis of its result Vehicle Number is found in String format. The system is implemented on hardware with the integration of Matlab, and its performance is tested on real images & a working prototype is developed. It is observed from the experiments that the developed system successfully detects and recognizes the vehicle number plate on real images of Pakistani vehicles.

Block diagram:

Components used:

·         Microcontroller

·         Web camera

·         Servo motor

·         Vehicle

·         Laptop

Conclusion:

The system is implemented for the automatic toll collection, beside this AVNR method developed in this research project can be also used for various other important applications such as, smart parking systems, access control, border control, finding stolen cars, airport parking & can also be used as a marketing tool. The design also deals with the issues of Number Plate Locating & Plate segmentation to make an efficient system. This research project facilitates the users in numerous ways like fair toll tax payment, automatic & easy way of paying toll tax.