NATIONAL SCHOOL of SURVEYING
DIVISION of SCIENCE
The National School of Surveying seeks to hire a Postdoctoral Fellow to provide research and development support to the Matariki Project https://www.otago.ac.nz/surveying/potree/pub/mrc/projects/matariki
This project will develop a novel way to make 3D-change maps of the Earth’s surface for monitoring of environmental resources and hazards.
The National School of Surveying at the University of Otago conducts world-class research that turns geospatial data into solutions for local, regional and global challenges. The University of Otago is one of New Zealand’s largest and most research-intensive universities, currently in the top 1% in the QS World University Rankings. It is a leader across all fields of academic endeavor, regularly topping the New Zealand Tertiary Education Commission’s performance indicators. The University is located in the vibrant southern city of Dunedin, which is a gateway to the beautiful Otago region of New Zealand.
This project will combine rigorous surveying methods and advanced computer-vision to automate high-resolution topographic mapping and change-detection from stereo satellite imagery, revealing subtle changes in the Earth’s surface over time. The successful applicant will work closely with experts in photogrammetry, computer vision, and high-performance computing to develop scalable and production-ready software for automated 3D-change detection. In this role, you will be responsible for prototyping, scaling-up, and validating a new automated image registration approach between satellite images.
You will survey the latest literature in computer vision and 2D/3D image/model registration, and explore new, creative ideas for feature detection and matching. Though iterative prototyping, testing, and validation, you will evaluate the optimal approaches for image registration and stereo-model triangulation. Your experience in efficient processing of large images will enable robust and rapid prototyping. Finally, you will work with HPC experts to accelerate algorithms and create production-ready code that can efficiently handle large images at scale.
Your skills and experience
The successful candidate will be strongly motivated to conduct research within the disciplines of machine vision and surveying.
You will also be expected to have:
• A PhD or equivalent doctorate in appropriate field such as imaging science, computer vision, remote sensing, photogrammetry, computer science, or software engineering.
• Published research results in national and international peer-reviewed journals.
• Past hands-on algorithm and software development in image science, environmental, geospatial, remote sensing, photogrammetry, or compute vision projects is required.
• Knowledge within more than half of the following subjects: (1) computer vision, (2) image registration, (3) photogrammetry, (4) 3D scene reconstruction, (5) object recognition, (6) signal processing, (7) machine learning, (8) feature detection, (9) uncertainty propagation, (10) localisation and mapping, (11) mathematical optimization.
This is a full-time, fixed-term position for two years available from early September 2020. The start date will be negotiable with the successful candidate.
The University of Otago is well-known for its collegiality and, combined with living in the vibrant city of Dunedin, provides opportunities for an excellent work-life balance. The Dunedin campus is ranked as one of the top 15 most beautiful in the world. The city has affordable living, excellent schools, and a great array of activities, ranging from Art and Literature to outdoor activities.
Candidates are requested to submit the following:
• A CV.
• A cover letter.
• Three referee contact details.
Applications quoting reference number 2001153 will close on Thursday, 20 August 2020.
To see a full job description and to apply online go to: https://otago.taleo.net/careersection/2/jobdetail.ftl?job=2001153
Equal opportunity in employment is University policy.
E tautoko ana Te Whare Wananga o Otago i te kaupapa whakaorite whiwhinga mahi.