The unpredictability and blockage of current transportation frameworks frequently produce traffic circumstances that endanger the security of the individuals and it can cause several problems in our life. To identify traffic offenders, we propose a system that will intelligently use image processing and deep learning to penalize them by recognizing their license plate. This will identify criminals, case of accident and initiates an emergency call on detecting it. A GPS-enabled traffic surveillance camera can detect the location of an accident and send a message or call the traffic control room with the accident's location. Retinanet, an object detection technique is used here for detecting vehicles which violates the traffic rule and its number plate will be localized by using image processing operation. The number plate information extracted by Optical Character Recognition and the car’s image are further processed by CNN for identifying criminals and alerts are given to the authorities. The accidents can be identified using deep learning and computer vision followed by delivering the signal to Arduino. The dislocated parts of object were detected from the images and fix it as an accident by CNN (Convolutional Neural Network) for certain threshold. When accident happens, alert is given for an emergency vehicle. This framework is made by fundamental GSM module. The system provides a robust and efficient method to detect red light offenders and criminals and provides an immediate emergency calling facility.
Traffic Surveillance System and Criminal Detection Using Image Processing and Deep Learning
2022-08-25
1226917 byte
Conference paper
Electronic Resource
English
Traffic surveillance by using image processing
IEEE | 2018
|Image Processing based Traffic Surveillance
IEEE | 2024
|Traffic surveillance using video image detection systems
Tema Archive | 2007
|Digital Image Processing Applied to Traffic Surveillance
NTIS | 1984
|