MULTI-IMAGE SUPER RESOLUTION AND NOISE REMOVAL FOR EMBEDDED IMAGING

Multi-image fusion is a good solution for performing super-resolution and denoising images, especially for low-cost image capture systems. However, in embedded systems such as satellites, drones or smartphones, the lightweight
embedded processing can be problematic. The innovative process optimizes the distribution of the computational load between the light-embedded processing device and the remote processing solution with a more powerful server. It can also considerably reduce the quantity of data to be transmitted or increase the quality of the result obtained for an equal quantity of data transmitted.

Applications

• Satellite
• Aerial recognition (defense or civil)
• AgriTech
• Burst mode on smartphone
• Multi-spectral/thermal/infrared imaging
• Sports camera

Competitive advantages

• Minimizes computer processing in the embedded system
• Drastically reduces the amount of data to be transmitted
• Drastically increases the quality of the result obtained with equal amount of data transmitted

Intellectual property

French patent application (FR2006182) filed on June the 12th, 2021. International patent application (WO/2021/250356 A1) filed on June the 10th, 2021.

Keywords

Super-resolution and denoising - Multi-image fusion - Embedded imaging - Distributed computing