Typical camera sensors have many identical pixels arranged on a fixed regular grid. These evenly spaced photo elements thereby define the sampling pattern and the related maximum spatial resolution of the camera.
Aiming to enhance the spatial resolution and to minimize aliasing effects, it was found that using non-regular sampling patterns in combination with sophisticated reconstruction methods can increase the resolution per sampled pixel.
The reconstruction algorithms are required to resample the image to a regular grid, for disaplaying, storing or further image processing steps.
My research focuses on designing and investigating novel, non-regular sampling patterns for camera sensors as well as evaluating different resampling approaches, which include neural network reconstruction, as well as classical signal processing techniques.
