Real-Time Search and Rescue using Remotely Piloted Aircraft System with Frame Dropping

Usage of Artificial Intelligence (AI) technology to aid the Remotely Piloted Aircraft System (RPAS) helps to get accurate imagery along with vital ground details, which as a result boosts the Search and Rescue operations. Since the search must be done quickly, real-time video processing is essential for survival. Our solution attempts to integrate image processing, more specifically, the You Only Look Once (YOLO) algorithm to detect humans in all environmental conditions. Moreover, traditional methods of AI use Graphics Processing Units (GPU) instead of Central Processing Units (CPU). We solved the issue of low frame-per-second processing on the CPU with a newly designed frame-skipping algorithm. This improved method results in accurate and quick detection of humans and allows real-time detection.