License Plate Recognition
Overview
The increasing number of vehicles on roads globally has led to a surge in traffic congestion, resulting in economic losses and time wastage. The Mapon survey reveals that the European Union alone loses approximately 1% of its GDP due to traffic congestion annually, while in the United States, traffic jams lead to a delay of 5.7 billion person-hours each year. To tackle this issue, governments across the globe are exploring various strategies, and Automatic License Plate Recognition (ALPR) is one such technique that aims to automate traffic management and surveillance activities.
This project will explore the potential of ALPR technology in reducing traffic congestion and improving traffic management by leveraging its capabilities, including real-time monitoring of traffic, detecting traffic violations, and identifying stolen vehicles. The project will also evaluate the effectiveness of the technology in different scenarios and the associated costs to implement and maintain the system. The outcome of this project will provide insights into the feasibility and effectiveness of ALPR in traffic management, which can help governments and transportation authorities make informed decisions about adopting this technology.
The idea behind creation of an algorithm for License Plate Recognition is to minimize the costs spent on management of traffic that includes manual toll collection, monitoring of traffic on roads, public parking spots and residential or commercials gated communities. A license plate is at some level a unique identifier of a vehicle. In countries that do not support a distinct license plate requirement, the scanner can be replaced with one that scans the RFID from a vehicle’s front face. The overall purpose is to ease every process involved in a traffic management scenario. This distinction of this algorithm is that it works on live video feed as well as on images, stores data securely and shows useful analytics, with all of this information it still performs better than other applications that are currently in the market both in terms of speed and accuracy.
Collaborators
Contributions to this project have been made by a team of three, details mentioned as below:
Spandita Sarmah: UX Product Design, Algorithmic Implementation and User Testing
Rahulraj Singh: Application System Design and study of on-going research
Dr. M. Uma: Guidance on Pattern Recognition Algorithm
Duration
3 months
Key Features
The application does not deviate from its core usage – assisting users in automating the process of vehicular management by capturing license plates and running it on databases to provide the required insights for analysis and records retrieval. Some key features that the application helps with are listed below:
Automatic Identification of License Plates from Video Feed / Image Input
Recognition of License Plates
Storing all recognized Number Plates in a secure Data Warehouse
Search operations for retrieving License Plates stored in the past
Integration with Government provided databases to retrieve information about recognized License Plates
Statistical Analytics on Stored Plate information
The idea for UX mock here is to incorporate most of these features in a dynamic one-page application in order to make it easy to navigate for the end-user.
Product Highlights & Design Theory
Key highlights of the application include a single dashboard that operates as a desktop interface and enables all functionalities of the product. So the end user does not need any navigation for trying the basic features of the application.
One-Screen All-Functionality Ideology – Keeping the end user in mind, I derived to a conclusion that traffic or surveillance monitoring personnel work with the application for the entire day and encounter hundreds of vehicles. In that scenario, keeping core functionality in separate locations of the product will cause delay in operation. Therefore, the idea is to bring all core functionalities for a day’s work on one screen so there is no navigation required. Post the work done, export operations and reports can be viewed once in a week, fortnight or monthly.
The UX Mock-up is not a key aspect for the project and is just an ideation of how a surveillance based application for desktop can be designed. The core of the research is the functionality of the application.
The program begins with an input image that is extracted from the video feed. This image is converted to a binary image and then preprocessing and noise reduction is performed.
The next steps are around detecting the rectangular region that covers the license plate and then detects the numerical values under the extracted region.
Another unique feature is addition of these extracted values to a database hosted on the cloud. This database can be used to retrieve information from the databases back to the user when requested. This features allows investigations at a later time.
Algorithmic Implementation
The license plate recognition uses advanced image processing on the MATLAB platform to extract numbers from a static image or a live video feed.
Development Outcome
A detailed explanation of the algorithm can be observed in the research paper. In the section below I have mentioned a few UX mocks that are in the works.
Product User Interface
The main interface page is created with two page layouts keeping the recent experience trends in mind – A white contrast page and a dark contrast page for better night visibility and usage. The features on the home screen are simple for the intended user base.
After developing the prototype, we performed user testing with 10 participants
User Testing
Creating Personas:
We started off by creating a persona and empathy maps to understand our users better.
Name: Sarah Jones
Age: 35
Occupation: Software Engineer
Education: Bachelor's degree in Computer Science
Technology: Comfortable with using computers and mobile devices
Goals: Sarah is a busy working professional who wants to save time and avoid the hassle of finding parking spaces in crowded areas. She values convenience, efficiency, and accuracy.
Pain Points: Sarah is often frustrated with the amount of time it takes to find a parking spot in busy areas. She also finds it inconvenient to carry around physical parking permits or tickets. Sarah values her privacy and is concerned about the security of her personal information.
User Scenario: Sarah is on her way to a client meeting in a crowded downtown area. She drives to the client's office building and sees that the parking lot is full. Instead of wasting time searching for a parking spot, she uses the license plate recognition system to find a nearby parking lot with available spots. She registers her vehicle information and enters the parking lot. The license plate recognition system scans her license plate and records her parking time. After the meeting, she pays for her parking fee using the system and leaves the parking lot, saving time and avoiding the hassle of finding a parking spot in a crowded area.
Empathy Map
What Sarah says:
"I hate wasting time looking for parking spots."
"I wish there was a way to find parking spots quickly and easily."
What Sarah does:
Searches online for parking options before driving to her destination
Drives around looking for a parking spot, often feeling frustrated
What Sarah thinks:
"Why can't parking be easier and more convenient?"
"I don't want to waste time and energy searching for parking."
What Sarah feels:
Frustrated when she can't find a parking spot quickly
Annoyed when she has to carry around physical parking permits or tickets
Methodology:
Ten participants were recruited for this study. The study used a mixed-methods approach, including both qualitative and quantitative data collection methods. Participants were asked to complete three tasks: register their vehicle information, enter a parking lot and have their license plate scanned, and check their parking status and pay the parking fee (if applicable). Participants' interactions with the system were recorded using screen capture software. After completing the tasks, participants completed a survey and participated in a semi-structured interview to gather feedback on their experience using the system.
Results:
The qualitative data analysis revealed that participants found the system easy to use and navigate. However, some participants reported difficulty in registering their vehicle information, specifically with the license plate field. Participants suggested adding a format guide for the license plate field to make it easier for them to enter the information correctly.
The quantitative data analysis showed that the majority of participants (90%) were able to complete the tasks within the expected time frame. The system's overall usability score was 85.6, indicating good usability.
Discussion:
The results of the study suggest that the license plate recognition system's user interface screens developed by the project team are effective and easy to use. However, some improvements can be made to enhance the system's usability, such as adding a format guide for the license plate field.
Conclusion:
The UX research study provides valuable insights into the usability and effectiveness of the license plate recognition system's user interface screens. The study's findings can be used to improve the system's user experience and ensure that it meets the needs of its users.