Using cameras beyond photography
As the Canada Research Chair in Computer Vision, Michael Brown is looking for new ways to understand what camera images can tell us about the physical environment around us. With over one billion smartphones with embedded cameras sold yearly, a large portion of the world’s population now has a sophisticated imaging sensor at their disposal. These cameras are not only used to produce photographs, but are also increasingly used for tasks of a more quantitative, scientific nature, from skin cancer detection to crop disease monitoring to aiding autonomous driving.
The goal of most consumer camera manufacturers is to create visually pleasing photographs, and not necessarily to faithfully capture the imaged scene. As a result, digital camera processing pipelines apply significant proprietary image interpretation to generate images—different cameras capturing the same scene will produce different results. While this is ideal for photography, such image manipulation hinders the ability to use cameras for purposes beyond photography.
Brown has led numerous efforts to improve the reliability of consumer cameras for scientific applications. His research focus as the Canada Research Chair is to design new camera processing channels that will produce conventional photographic images as well as images suitable for scientific applications. This research program will help shape the future design of consumer cameras, and expand the range of applications in which they are used.