Automatic Traffic E Challan Generation using Deep learning

  • M. Suma Sree, N. Varalakshmi


People's careless and reckless attitudes are causing a severe issue with traffic law infractions, which is weakening society's moral foundation. The human aspect in our existing system continues to be a burden and yields subpar outcomes when it might have produced far better results, despite the fact that our country's traffic rules have significantly improved over the previous several years. The evolution of technology has ushered in a new era of intelligent urban mobility management. Our project aimed to develop a system which is cost efficient and accurate in detecting the vehicles which violate the traffic laws and regulations made by Indian government. Deploying Convolution Neural Network (CNN) model in Raspberry pi4 board to perform real time image processing operations and with the integration of Canny Edge Detection (CED) which enhances the system accuracy by removing noise and delineating vehicle boundaries in even blurred images. The output image of CED algorithm is sent to Optical Character Recognition (OCR) to extract the number plate using OpenCV library.

Finally Global System for Mobile Communication (GSM) module springs into action issuing instant Short Message Service (SMS) notification to authorities for prompt E Challan generation. The proposed system is able to achieve 96% accuracy rate to detect the license number plate.