The Autonomous search and rescue operations in hazardous en- vironments demands a robust perception, precise localization, and intelli- gent navigation. This paper presents an integrated robotic system on the TurtleBot3 Waffle platform combining vision based object detection with Si- multaneous Localization and Mapping (SLAM) for autonomous exploration and hazard identification. The system employs a dual detection strategy: YOLOv8 based fire and smoke detection trained on an augmented dataset of 3000 images (with metric values of mAP50: 0.84, Precision: 0.87, Re- call: 0.83) and HSV color segmentation for human victim identification, integrated with SLAM Toolbox for real time mapping and Nav2 for au- tonomous navigation. A frontier based exploration strategy enables system- atic area coverage without manual intervention. The experimental validation in Gazebo simulation achieves a centimeter level hazard localization accu- racy and 97% map coverage through autonomous exploration, demonstrating effectiveness in obstacle dense warehouse environments under varied visibil- ity conditions.
N et al. (Thu,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: