Isolates blood vessel from .tiff image stacks produced by photoacoustic scans. Was designed to allow accurate segmentation on noisy image sets. Works in 3 steps:
- Changes brightness & saturation in image to make blood vessels clearer.
- Runs a noising algorihm to remove random noise
- Uses edge detection & morphologicl funcitions to find the blood vessels
- Currently working on containerising this application using docker.
- Adding detection and deconvolution of a valid point spread function to remove motion blur. This was preciously attempted with a circular point spread function, however this did not improve the image quality
- Implementing watershed algorithm to improve vessel segmentation