Search and Rescue Operations with an Unmanned Helicopter

Abstract

We tested the performance of both electro-optical (EO) and long-wave infrared (LWIR) video sensors on an unmanned helicopter in a simulated Search and Rescue (SAR) setting. Our objectives were to examine whether humans and other objects associated with (SAR) operations could be reliably observed in the video sensor imagery in direct light and shade. Our unmanned helicopter flight occurred in a forest in the U.S. Pacific Northwest under a Certificate of Authorization (COA) issued by the Federal Aviation Association (FAA). Analysis of the EO and LWIR video imagery from several hover positions determined that many of the SAR objects could be reliably detected. Detection success was heavily influenced by factors such as distance, object size, material and color. In direct sun light, the EO sensor performed more reliably than the LWIR sensor in detecting most objects. The LWIR sensor, however, appeared to be preferable to the EO sensor when tracking human subjects in shady environments. We believe unmanned aircraft can offer SAR operations flexibility in reaching areas that might endanger manned flight crews. In addition, unmanned aircraft also have the capability to fly closer to the ground, and at slower speeds than some manned aircraft, which potentially leads to greater image resolution and ultimately increases detection success.