Fibre Bundle Image Processing/Core Removal (Matlab)

Removes core pattern from a fibre bundle image by triangular linear interpolation

Removes core pattern in fibre bundle images by triangular linear interpolation. Calibration requires identifying location of each core, and so all images must be clearly resolved. Method is slower than simple spatial filtering but can deal better with broken cores or more complex reconstruction schemes (super-resolution etc.). It also may work better for coherent imaging where simple filtering does not preserve integrated intensity for each core.

Flat fielding involves capturing an image of uniform target (or simply shining a light through he bundle). The intensity value of each core is extracted from this flat field image and used to scale intensity values from each subsequent image. This accounts for differences between core transmission and, more importantly, any imperfections in finding the centre of each core.

Two methods of extracting core values have been implemented. The point method simply takes the pixel value at the detected centre of the core. This tends to lead to variations in intensity across the image, depending on where the core happens to sit relative to the pixel grid of the camera. Flat field correction is therefore critical.