Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.698709
Title: Near real-time monitoring of buried oil pipeline right-of-way for third-party incursion
Author: Olawale, Babatunde Olumide
ISNI:       0000 0004 5992 5083
Awarding Body: University of Sussex
Current Institution: University of Sussex
Date of Award: 2016
Availability of Full Text:
Access through EThOS:
Full text unavailable from EThOS. Please try the link below.
Access through Institution:
Abstract:
Many security systems employing different methods have been proposed to protect buried oil pipelines transporting petroleum products from the well head via the refinery to: depots and other receiving stations. Currently there is a security gap in the monitoring of these buried pipelines in real time and in keeping them protected from third party interference. This thesis addresses the problem of monitoring these systems by developing an automated image analysis system with the aid of a low-cost multisensory Unmanned Aerial Vehicle (UAV) for monitoring of buried pipeline right-of-way (ROW). The method used in this research is based on the identification of threat objects of interest from the video frame sequences of the pipeline right-of-way acquired by the UAV. This is achieved by training the system to recognise objects of interest using trained correlation filters. To determine the geographical location of detected objects, the Video frame sequences captured by the UAV platform were ortho-rectified to form ortho-images which were then mosaicked to form a seamless Digital Surface Model (DSM) covering the test area using a photogrammetry model. The DSM formed from the mosaicking of ortho-images is then emerged with a digital globe for geo-referencing of detected objects. Experiments were carried out on a test field located in United Kingdom and Nigeria, where video and telemetry data were collected, then processed using the techniques created in this research. The results demonstrated that the developed correlation filter was able to detect objects of interest despite the distortions that come with the object image, due to the fact that the expected distortion was compensated for using the training images. When compared with the 6 control points in the digital globe the accuracy of the two-dimension DSM gave a misalignment error of between 2 and 3 metres.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.698709  DOI: Not available
Keywords: T0059.5 Automation
Share: