Forlenza, Lidia (2011) Vision based strategies for implementing Sense and Avoid capabilities onboard Unmanned Aerial Systems. [Tesi di dottorato] (Unpublished)
Download (5MB) | Preview
|Item Type:||Tesi di dottorato|
|Uncontrolled Keywords:||Unmanned Aerial Systems, Collision Avoidance, Electro-Optical Systems, Image Processing Algorithm, Multisensor Tracking.|
|Date Deposited:||08 Dec 2011 19:29|
|Last Modified:||30 Apr 2014 19:47|
Current research activities are worked out to develop fully autonomous unmanned platform systems, provided with Sense and Avoid technologies in order to achieve the access to the National Airspace System (NAS), flying with manned airplanes. The TECVOl project is set in this framework, aiming at developing an autonomous prototypal Unmanned Aerial Vehicle which performs Detect Sense and Avoid functionalities, by means of an integrated sensors package, composed by a pulsed radar and four electro-optical cameras, two visible and two Infra-Red. This project is carried out by the Italian Aerospace Research Center in collaboration with the Department of Aerospace Engineering of the University of Naples “Federico II”, which has been involved in the developing of the Obstacle Detection and IDentification system. Thus, this thesis concerns the image processing technique customized for the Sense and Avoid applications in the TECVOL project, where the EO system has an auxiliary role to radar, which is the main sensor. In particular, the panchromatic camera performs the aiding function of object detection, in order to increase accuracy and data rate performance of radar system. Therefore, the thesis describes the implemented steps to evaluate the most suitable panchromatic camera image processing technique for our applications, the test strategies adopted to study its performance and the analysis conducted to optimize it in terms of false alarms, missed detections and detection range. Finally, results from the tests will be explained, and they will demonstrate that the Electro-Optical sensor is beneficial to the overall Detect Sense and Avoid system; in fact it is able to improve upon it, in terms of object detection and tracking performance.
Actions (login required)