AI-BASED TRAFFIC SURVEILLANCE FOR VIOLATION DETECTION AND DENSITY ANALYSIS

  • Unique Paper ID: 174647
  • Volume: 11
  • Issue: 11
  • PageNo: 522-529
  • Abstract:
  • Traffic Offences such as over-speeding, not wearing helmets and seat belts, talking on mobiles, jumping through red lights, and lane jumping are key road safety hazards. The present system is heavily dependent on manpower-based surveillance and post-incidence CCTV data analysis, thereby creating delays, and errors due to a lack of manpower and also variable applications. Manual surveillance also takes more time and will make it infeasible for efficiently tracking delinquents. This paper suggests an artificial intelligence-driven innovative traffic rule violation detection system integrating YOLOv8 for real-time object detection and rule-violation classification, and an optical flow-based Speed Estimation mechanism for over-speeding detection. YOLOv8 detects violations like mobile phone usage, triple riding, red light jumping, lane deviation, and lack of helmets and seat belts. Optical flow-based approaches estimate vehicle speeds to detect over-speeding violations. Besides that, density analysis is incorporated to gauge traffic congestion as well as how it relates to rule breaches to enable more flexible enforcement. Leveraging computer vision, machine learning, and automation with AI support, the envisaged system enables real-time breach detection, best-in-class processing, minimizing human intervention, as well as safer roads. The solution fits with the smart city strategy by offering an adaptable, sturdy, and evidence-based solution for traffic law enforcement.

Cite This Article

  • ISSN: 2349-6002
  • Volume: 11
  • Issue: 11
  • PageNo: 522-529

AI-BASED TRAFFIC SURVEILLANCE FOR VIOLATION DETECTION AND DENSITY ANALYSIS

Related Articles