March 14, 2019, by Lexi Earl
Meet the Beacon: Dr Michael Pound
Michael Pound is a Nottingham Research Fellow in Computer Vision and member of the Future Food Beacon of Excellence. Mike is a computer scientist whose research focuses on the development of novel computer vision techniques. Within the Beacon, he is working with bioscientists to apply these computer vision techniques to plant phenotyping, providing accurate measurement of 2D and 3D shape information of plants. The bioscientists can use this information to explain the underlying mechanisms for how plants grow. Dr Pound has developed tools that allow biologists to measure plant root systems. These have been adopted by researchers all over the world.
In this interview, Mike talks about his work at the Future Food Beacon, and his research plans.
What is your role with the Future Food Beacon?
As a Nottingham Research Fellow, I will be developing and applying modern computer vision techniques to the challenging domain of plant phenotyping. If we are to understand plant development, we must be able to measure them quickly and accurately. I will be working on numerous projects within the Beacon, bringing modern computing power to bear on tasks that would otherwise take hundreds of hours for humans.
Why Nottingham and why the Future Food Beacon?
Nottingham is a world leader in plant phenotyping. The University has long-established collaborations in this area, and the Hounsfield Facility offers state-of-the-art facilities for root imaging in soil. I will be working to match these facilities with equally accurate shoot imaging techniques, both within glass-house and field environments.
How would you explain your research?
Despite recent advances, computer vision still lags some way behind the human visual system. We can compare and measure objects very quickly, usually from a single angle. Humans can even do this with extremely complex objects, such as plants. However, it’s not feasible for a person to sit and measure thousands of plants every day, we need ways to automate this. I work jointly with bioscientists to apply modern computer vision techniques to their challenging images, providing accurate measurement of 2D and 3D shape information of plants. This information feeds back into their research, where they use it to explain the underlying mechanisms of how plants grow.
What inspired you to pursue this area?
I find that working in a collaborative area such as this offers me the chance to address interesting problems with real-world impact. I also get to work with a variety of world-class researchers from many domains, which ensures that my work is always engaging and rewarding.
How will your research affect the average person?
Data from my imaging approaches will be used to inform plant breeding and genomic research. In turn, this research will have a huge impact on the ongoing problem of global food security. Developing bigger, healthier plants that can grow in challenging environments is a pressing issue, but one where we can have huge impact. Novel imaging approaches I develop can also be used outside of plant imaging, such as in medical image analysis, further benefitting the public.
What’s been the greatest moment of your career so far?
I’ve had a very rewarding career so far, throughout my PhD and time as a post-doc. Perhaps my favourite bit was when I released some new tools based on my algorithms, allowing biologists to quickly measure their root systems. Very quickly these tools were adopted by the research community, they have now seen thousands of downloads and are in use throughout the world.
How will being based at the University of Nottingham and joining Future Food help you achieve your goals?
Nottingham has world-class facilities to support my research. The Future Food Beacon will allow me to collaborate with a variety of researchers from different backgrounds, on new and exciting projects. The Beacon also has a strong mentoring program, which will allow me to establish myself as an independent researcher within the University.
What aspects of your research and role are you looking forward to?
I am very excited by the idea of using new deep learning approaches to drive automated camera equipment. Fixed cameras have limited our ability to collect the best images, I believe that we can transform image capture using new artificial intelligence methods. This is also a very interesting research question; can a computer learn to perform the same task as an expert human operator?
The Future Food Beacon will also support a very wide variety of projects, I will be able to attend workshops and foster collaborations with more researchers, applying for funding and building a team to take on these interesting and important challenges.