PhD Studentship in Extracting 3D Objects from Point Clouds

Lancaster University

Atlantic Geomatics

Closing Date: This position is open until filled, early submissions are encouraged.

Stipend: A studentship is for 36 months with a £15,000 per annum stipend for UK/EU students.

Topic

A fully funded 3 year PhD Studentship is available to work on the automatic generation of 3D objects from point clouds.

3D point clouds are generated by sensing systems such as terrestrial and airborne laser scanning systems. The massive point datasets generated by such systems represent surface features by the spatial position of points in 3D only. The point data are an indirect representation of the surface features of interest such that identification and characterisation of those surface features is an inverse process for which there are many possible solutions. Surface features of interest may include buildings and their type, vehicles and their type, trees and their functional type and so on. A particular motivating application is the automatic extraction, labelling and positioning of gravestones from laser scanning data, which has the potential to augment existing aspatial records.

The aim of this project is to research, design and implement a solution for the automatic extraction of 3D objects with application testing in several use cases. The developed system will identify objects, label them and position them in an inter-operable database as isolated object entities that can be handled in 3D by an operator (e.g., moved around, switched on and off etc.). This novel processing, thus, has the potential to greatly amplify the useability of digital point cloud data, with very wide applicability.

Requirements

Candidates should have a 1st class degree (or equivalent) in computer science, statistics or geography (with a specialism in a spatial quantitative area, e.g., remote sensing or GIScience). The candidate should have interests in spatial data handling and ideally in the acquisition and processing of point cloud data through systems such as ground and airborne LiDAR. Good programming skills is an essential requirement. Experience in the processing of large sets of remotely sensed data is an advantage, although not required.

This project will require a self-motivated and high performing candidate. The expectation is for a candidate with excellent interpersonal skills who will perform research work of high quality and publish their work in leading internationally refereed journals. This doctoral research project will be supervised by Lancaster University and Atlantic Geomatics and is part funded by the European Regional Development Fund.

Application details

To apply for this opportunity please apply by email to fst-pg-admissions@lancaster.ac.uk with the subject “3D Object Extraction” and include:

  • A CV (2 pages maximum)
  • Cover letter
  • University grade transcripts

Lancaster University

Lancaster University is among the best in the UK. Top 10 in all three major national league tables, we’re also highly ranked in international league tables such as the QS world rankings. We have invested over £450m into our campus in the last ten years, including our new Engineering building, the refurbishment of the Library, Physics and Chemistry facilities and a £20m sports centre and facilities, with a further £135m investment planned. Lancaster is the best in the North West for graduate prospects, according to The Good University Guide and one of the top universities in the UK for overall student satisfaction according to The National Student Survey.

Lancaster University’s Data Science Institute is creating a world-class data science research capability that sets the global standard for a truly interdisciplinary approach to contemporary data-driven research challenges.

Contact Us

For informal enquries applicants should contact Professor Peter Atkinson.