APPLYING MACHINE LEARNING AND OBJECT DETECTION FORIDENTIFYING TRAFFIC RULE VIOLATION: OVER SPEEDING

Abstract

We have observed that traffic offences are manually inspected by the traffic department in many developing nations, such as India. Such systems enable traffic management and traffic law enforcement. Following this, the present study focuses on Traffic violations as one of the key issues hindering the efficient operations and safety aspects in the urban traffic area. Currently, traffic violations are controlled using enforcement measure that demands a high manual workforce, and resource which entails high economic cost. Contrary, the use of automatic traffic violation detection systems in many of the urban cities of the developed and developing countries is being implemented for a higher degree of traffic enforcement and adherence to the traffic operation rules. Following this the present study focus on the applicationof machine learning for automatic traffic violation detection for weak- lane disciplined mixed traffic conditions. The resulting exhibit that the developed model can detect traffic violations of over-speeding at an accuracy of 95%. The traditional methods like speed camera use the basic principle of the Doppler Effect and RADAR technologies but at present,Due to their drawbacks which vary from the expense of purchasing equipment, increase in work and labour costs, and lack of monitoring, they are ineffective. As a result, it is a straightforward, affordable, and practical speed detecting method in Mixed traffic conditions. So here, we prefer automatic speed violation detection. The suggested solution will be highly accurate and cost-efficient.