This thesis presents the development and implementation of a software framework for estimating the position of a drone during flight. This framework is based on an algorithm known as the Unscented Kalman Filter (UKF), a recursive method of estimating the state of a highly nonlinear system, such as an aircraft. In this thesis, we present a UKF formulation specially designed for a quadcopter carrying an Inertial Measurement Unit (IMU) and a downward-facing camera. The UKF fuses data from each of these sensors to track the position of the quadcopter over time. This work supports a number of similar efforts in the robotics and aerospace communities to navigate in GPS-denied environments with minimal hardware and minimal computational complexity. The software framework explored in this thesis provides a means for roboticists to easily implement similar UKF-based state estimators for a wide variety of systems, including surface vessels, undersea vehicles, and automobiles. We test the system's effectiveness by comparing its position estimates to those of a commercial motion capture system and then discuss possible applications. / Master of Science