Multimodal Ambulance Detection Using Image Processing and RF433MHz Sensor

Abstract

Fast and unobstructed movement is a critical issue for ambulances in emergency medical services. This project aims at the development of an image processing and RF433MHz sensors-based multimodal detection system for ambulances to enhance traffic management systems. The proposed system in- tegrates computer vision techniques to differentiate ambulances from raw surveillance camera feeds and an RF433MHz sensor module that detects the RF signals given off by ambulances equipped with RF transmitters.

The image processing module utilizes object detection and feature recognition algorithms to identify the features of am- bulances, including sirens and vehicle markings. The RF sensor will simultaneously detect ambulances in a certain distance to ensure redundancy and robustness in detection. Once detected, the system will communicate with the traffic signal controllers to prioritize ambulance movement by dynamically controlling traffic lights.

Ambulance detection under several scenarios, especially poor visibility and heavy traffic load conditions, are reinforced and ensured to have increased reliability and accuracy through a multimodal system. The scalable system is therefore made cost- efficient with easy integrations into an already existing city infrastructure for efficient traffic management for decreased response time and improved quality emergency medical care services.