Material-Object Detection in Hyperspectral Images using Machine-Learning
Engineering, IT, Mathematics and Statistics, Medical, Biological and other Sciences
- Due to the sensitivity and security of this project, students must have Australian Citizenship to apply. Any applicants not meeting this requirement will automatically be deemed ineligible for this project.
Hyperspectral images have spectral information that enables material classification using the unique spectral characteristics of materials. This project aims to explore extending the demonstrated capability of machine-learning (ML) to detect and recognise objects based on spatial information to use the additional spectral-information in hyperspectral images to perform material-object detection.
RESEARCH TO BE CONDUCTED
Develop and evaluate the performance of a ML system to perform material-object classification/detection (e.g. red- plastic-square-tile versus red-clay-square-tile) in image-data from imaging hyperspectral sensor. The project will include collection of a labelled image-dataset using a ground-based imaging hyperspectral system.
Using transfer learning and/or fine-tuning on a pre-trained net to detect a material-object target in airborne hyperspectral data.
Develop and evaluate the performance of a ML system to regress the atmospheric transmission, upwelling and downwelling (TUD) spectrum from hyperspectral image data and perform atmospheric correction to retrieve ground leaving reflectance.
SKILLS WISH LIST
- Python (for Deep-learning/machine-learning. E.g. Keras, PyTorch, TensorFlow2)
- Lab/Field data-collection skills
- Desirable/Advantageous: VNIR/SWIR/LWIR spectroscopy, chemistry, physics, computer-vision
Proof-of-concept for material-object detection in hyperspectral image data using ML. Initial baseline for performance. Identification of hurdles or blockers to full development of ML material-object detection system.
The intern will receive $3,000 per month of the internship, usually in the form of stipend payments.
It is expected that the intern will primarily undertake this research project during regular business hours and maintain contact with their academic mentor throughout the internship either through face-to-face or phone meetings as appropriate.
The intern and their academic mentor will have the opportunity to negotiate the project’s scope, milestones and timeline during the project planning stage.
CONNECT WITH APR.INTERN