Intelligent Infrared Imaging Sensors Using Machine Learning
Machine vision systems using visible cameras are used for inspection and automation in the industrial and defence sectors with a market value of $14 billion USD.
Learning algorithms can be trained to automatically detect features and even provide depth perception from 2-dimensional video data.
However, for many applications, infrared and thermal vision can see features and see-through obscurants where visible cameras cannot. Seeing through rain, smoke, fog, imaging the temperature of an object or the water content of food and crops are all scenarios where infrared vision outperforms visible.
The overall direction of this research program is to apply machine-learning techniques to emerging infrared imaging systems to solve current problems in industry.
One project is to demonstrate passive ranging and augmented reality that can be used for military scenarios on land and sea. A second project is to investigate smart thermal imaging to assist medical operations. There will be interaction with local companies including BAE Systems.
This project will involve the assembly and alignment of optical cameras and laser radars, recording video data and developing computer algorithms to demonstrate image enhancements. The student can have interest in either the infrared imaging sensor or the machine learning algorithm components or both.
How do we feed the world’s growing population? How do we save our wildlife from extinction? Got an idea that will build a brighter, greener world?
Australian high school students are invited to submit a short video about one of Australia’s big science challenges.