This software was developed in order to automatically recognize license plate numbers and detect if the odd even rule is followed and inform the defaulters automatically with a system generated email so that manual intervention is virtually eliminated
The odd-even traffic rule states that on certain days of the week only vehicles with license plate numbers ending with an odd digit can use the road and only even (last digit) numbered license plate vehicles on the other days. This reduces air pollution by vehicular emmission and traffic congestion upto 50% on a daily basis. Manually checking for vehicular number plates and determining for rule compliance and later notifying the defaulters is a tedious and error prone process. This software provides the complete automation from plate number recognition on the images received upto querying the RTO database for defaulter licensee credentials and using the extracted credentials to notify the defaulter of the necessary details completely eliminating human interference.
The software developed receives images of license plate numbers(already captured) from an interceptor camera and uses computer vision to process these images and extract the license plate number as a string. A sample database which is made for illustrative purposes contains the license number and the licensee's corresponding contact details. Once the CV algorithm interprets the license number, the database is queried for corresponding contact details/credentials using the extracted license number as a key. These details are then fed into another system which generates an email and message and uses SMTP to send emails and messages to the defaulters whose ID's and contact numbers were obtained from the database.
Completely developed in python, for license plate recognition, adaptive threshold and bilateral filter preprocessing was done and pytesseract was used to perform the OCR and read out the license plate number. Then the last digit is checked for consistency with the odd even rule. mysql.connector was used to query the database with the necessary commands to extract the email-id and contact number. Then a system generated email and message is sent to the corresponding ID's through smtp and twilio API to send mails and messages to the defaulters. A sample front-end is also provided for reference.
Install all dependencies. All files,images must be in the same and single folder. mail.py file contains the entire process from OCR to mail sending, it randomly picks the number of vehicles on the road and iterates through the images of the vehicles, a list is stored with names of images for all of the randomly generated index numbers. the mys.py file contains the code to create the database or table. and index.html isthe sample front-end. A few images are also given for reference and verification of the OCR accuracy. Feel free to tweak anywhere to suit your needs.