This research combines multimodal large language models with instance segmentation techniques to create intelligent traffic monitoring systems that comprehend complex urban conditions and provide intuitive user interfaces.

• Enables real-time adaptive traffic control through advanced visual analysis
• Creates systems that can interpret and respond to dynamic traffic conditions
• Enhances road safety through improved monitoring and collision detection
• Supports smart city initiatives by optimizing traffic flow and reducing congestion

This engineering innovation represents a significant advancement for Intelligent Transportation Systems (ITS), offering municipalities more efficient ways to manage traffic infrastructure while improving public safety outcomes.

Leveraging Multimodal-LLMs Assisted by Instance Segmentation for Intelligent Traffic Monitoring