Date of Award


Access Type

Thesis - Open Access

Degree Name

Master of Science in Aerospace Engineering


Graduate Studies

Committee Chair

Dr. Richard Prazenica

First Committee Member

Dr. Hever Moncayo

Second Committee Member

Dr. Troy Henderson


This thesis presents the development of vision-based state estimation algorithms to enable a quadcopter UAV to navigate and explore a previously unknown GPS denied environment. These state estimation algorithms are based on tracked Speeded-Up Robust Features (SURF) points and the homography relationship that relates the camera motion to the locations of tracked planar feature points in the image plane. An extended Kalman filter implementation is developed to perform sensor fusion using measurements from an onboard inertial measurement unit (accelerometers and rate gyros) with vision-based measurements derived from the homography relationship. Therefore, the measurement update in the filter requires the processing of images from a monocular camera to detect and track planar feature points followed by the computation of homography parameters. The state estimation algorithms are designed to be independent of GPS since GPS can be unreliable or unavailable in many operational environments of interest such as urban environments. The state estimation algorithms are implemented using simulated data from a quadcopter UAV and then tested using post processed video and IMU data from flights of an autonomous quadcopter. The homography-based state estimation algorithm was effective, but accumulates drift errors over time due to the relativistic homography measurement of position.