Body Detection in CSSR Missions: Smart Rescue Operation using Robot for Enhanced Body Detection

Abstract

Collapsed Structure Search and Rescue missions are essential during disasters such as earthquakes, explosions, and structural collapses, where rapid victim identification can significantly influence survival outcomes. However, locating trapped individuals within unstable debris environments remains challenging due to restricted access, low visibility, and safety risks to rescue personnel. This study presents a smart robotic rescue system developed to enhance human body detection in such hazardous conditions. The proposed system integrates a mobile robotic platform with a vision-based detection framework capable of identifying human presence in real time. A deep learning–driven object detection model is employed to analyze visual data captured from complex and obstructed environments. The system was evaluated using a dataset designed to simulate collapsed structure scenarios with varying lighting and obstruction levels. Detection performance was assessed using precision, recall, and overall accuracy measures. Experimental findings indicate that the proposed approach achieves reliable detection performance while maintaining operational efficiency suitable for time-sensitive rescue missions. By reducing direct human exposure to dangerous zones and improving victim localization speed, the system supports safer and more effective rescue operations. This work contributes to the advancement of intelligent robotic solutions for disaster response and emergency management.