Video-based autonomous navigation

As a part of a precision agriculture project OKSI operates a suite of remote sensing sensors including multi-, hyperspectral, thermal IR, and a color video, from a Cessna aircraft. In order to properly register and georeference the imagery OKSI hasdeveloped algorithmic tools for image registration to a landmark reference (e.g., satellite) image, and for platform attitude (roll pitch and yaw extraction). These tools can also address differences in brightness or illumination and seasonal changes inthe scene.In Phase-I we will use our sensors to acquire flight data under various environment conditions including day/night and different terrain types. We will use satellite imagery from OKSI's archives to test the robustness of the algorithms in imagegeoregistration, and platform attitude determination. A miniaturized low cost flight sensor package will then be designed for use on board a UAV down to the size of SWARM.The experience gained in recent years through OKSI's precision agriculture project will be used to develop "autonomous video based navigation system" for use under denial of GPS conditions. Similarly, enhancements in the technology will be used by OKSI tofurther its remote sensing in precision farming business opportunities.