[robotics-worldwide] [jobs] Research Associate Post - Visual NDT/Metrology (188409)
aamir.khan.101 at strath.ac.uk
Wed Feb 6 03:21:58 PST 2019
The Department of Electronic and Electrical Engineering (EEE) at the University of Strathclyde is seeking to recruit a Research Associate to participate in a multi-partner collaborative research project focused on automated visual inspection / Non-Destructive Testing. Initially, this is for a 14-month post, but we anticipated this will be extended as future funding is secured.
The role will involve working with a research team to develop a novel software solution to pipe scanning. It will involve considerable lab work, designing experiments, collecting data and determining new methods to determine data quality. You will consider parameters such as lighting, focus, motion blur and data compression. You will develop novel algorithms that improve our capability to produce ‘metrology grade’ 3D models of the inside of industrial pipework. You will implement these algorithms into a prototype and quantify performance. You will interface directly with industrial partners to help transfer new technology into industry for exploitation.
This project will feed into several areas of state-of-the-art research in photogrammetry and sensor fusion running at the University supported by a highly active and growing team of talented researchers. Ongoing projects involve 3D photogrammetry deployed in a diverse and challenging range of applications including the visual inspection of wind turbine blades from drones, 3D model generation to aid medical images of breast tissue and a system for creating photorealistic models of engines. You will have access to world class equipment in the Facility for Innovation in Structural Testing (FIRST) Laboratory, in the Technology Innovation Centre (TIC) building at the University of Strathclyde.
This specific role will consider automated 3D photogrammetry for pipe inspection. The research will use state of the art pipe inspection hardware to develop software and algorithms to map the inside of pipework. You will use a form of ‘structure from motion’ to generate a textured 3D model of the sample which will be used to guide the inspection process.
1. 3D model reconstruction, particularly in situations with low surface detail.
2. Camera calibration and optimization of camera parameters/lighting.
3. Novel algorithm development
4. Fusion with other datasets including laser scans and inertial measurement units
This is a fertile area for publication and it is expected that developments are published in high quality journals. The role will require the development of robust software to process real world data and also the design and implementation of laboratory experiments. You will work as part of a research team of 15 members, featuring academic staff, postdoctoral researchers and PhD students interested in visual processing research.
To be considered for the role, you will be educated to a minimum of PhD level in a relevant subject (e.g. Electrical or Software Engineering); or have significant relevant experience in addition to a relevant degree. You will have demonstrable capability in algorithm development in Matlab or equivalent. You will have a strong background in algorithm development, linear algebra and optimisation. You will have an ability to conduct individual research work, to disseminate results and to prepare research proposals. You will have a proven track record of high-quality publications. You will have excellent written and verbal communication skills, with an ability to listen, engage and persuade and to present complex information in an accessible way to a range of audiences. You will have the ability to work as part of a team, collaborating effectively with both academic and industrial partners.
For initial enquiries, please contact Dr Gordon Dobie, Senior Lecturer, gordon.dobie at strath.ac.uk and/or Dr Aamir Khan, Research Associate, aamir.khan.101 at strath.ac.uk.
Dr Aamir Khan,
University of Strathclyde, UK
More information about the robotics-worldwide