Design of Zebra Cross Violation Detection Model in Traffic Light Using the Adaptive Background Subtraction Method

Abstract

Traffic violations often occur in India, both by law and non-law enforcement agencies. The destruction occurred due to lack of public awareness for orderly traffic. So it is considered normal and become a habit if allowed to continue. Various categories of violations occurred such as running a red light, not using a helmet, not turning on a vehicle's light, not carrying a driving certificate, going against the flow, and crossing the crossing markers. Not infrequently these violations can harm other road users and result in accidents. One type of traffic violations that often results in accidents in traffic is the driver breaking or breaking into traffic light when the red conditions are on. In this research, zebra cross detection system is created using the Adaptive Background Subtraction method using Raspberry Pi3. The adaptive method of development plan planning is used to keep track of the images that are in circulation for capturing the object in traffic light using a camera; the image obtained can capture objects that violate the detection line. Then the results of the object that was captured when breaking will be given a caution in the form of a buzzer. This system offers high precision in identifying the vehicle that commits the infraction, helping to improve the safety of pedestrian and vehicular traffic, and providing municipalities and cities with a new source of income.

Keywords: Violation, Adaptive Background Subtraction, Traffic Light, Raspberry Pi3.