Grant: Discovery Grants Program - Individual

Hyperspectral (HS) imaging can provide much more information than any other types of optical imaging, such as panchromatic (Pan), multispectral (MS), and super-spectral (SS) imaging, because it records more than 100 narrow spectral bands to capture direct spectral identifications of surface materials. Therefore, HS imaging has a wide range of applications, including remote sensing (agriculture, forestry, environment, geology, ocean, atmosphere, and climate change), and defence and security. However, due to the four crucial limitations of HS imaging, i.e. (a) limited spatial resolution, (b) low signal-to-noise ratio, (c) narrow ground coverage, and (d) huge data volume, its application scope has been significantly limited. Canada is planning to launch a space-based HS mission-WaterSat. But, due to the four crucial limitations, its spatial resolution is limited to 100m, significantly limiting the scope of its applications. In 2020, I obtained a research grant from the Canadian Space Agency (CSA), to conduct Experimental Research to develop and validate a groundbreaking Sensor-Specific, Tri-level and Tri-sensor Fusion (2Tri-Fusion) solution to overcome the four crucial limitations of WaterSat. This Discovery research program will conduct Theoretical Research and In-depth Image Matching Research, to develop and validate a groundbreaking Universal 2Tri-Fusion solution to significantly overcome the four crucial limitations of all HS imaging systems onboard all platforms, including satellite, airplane, ground vehicle and hand-held device platforms.   �The 2Tri-Fusion approach will be developed based on the principle of human vision, compressed sensing, system of systems, and our world-leading research in bi-level sensor fusion.   �The Theoretical Research will aim to: (1) understand the phenomena of the information transition in the 2Tri-Fusion process; (2) develop a universal mathematical model for a Universal 2Tri-Fusion for all HS sensors; and, (3) validate the Universal 2Tri-Fusion using available data from all HS sensor systems.   �The In-depth Image Matching Research will develop a new solution to overcome one of the most crucial problems in close-range multi-sensor imaging systems, i.e. misregistration between multi-sensor images taken at close distances, in order to extend our 2Tri-Fusion solution from space-based HS applications, to air- and ground-based HS applications. The successful outcomes of the Theoretical Research will not only create new knowledge in multi-sensor fusion, but also lead to a breakthrough in space-based HS imaging, making a significant contribution to Canadian and international future space HS missions. The successful outcomes of the In-depth Image Matching Research will lead to a breakthrough in the co-registration of close-range multi-sensor images, making a significant contribution not only to air- and ground-based HS imaging, but also to other multi-sensor imaging systems (such as multi-camera smartphones).