Description

Biomedical images are often corrupted by a large amount of noise which makes their analysis difficult. Smoothing is an accepted methodology to reduce the effect of noise. However, normal smoothing blurs the bounderies, particulary in images containing fine structures. Hence, in this project we attempt to perform anisotropic smoothing that preserves the bounderies while reducing the noise.

The user can adjust some parameters depending of the diffusion type:
1) Iteration: Quantity of loops wanted.
2) Threshold: Value of the gradient norm from which the operation is done.
3) Delta: Ponderation of the final increment.
4) Sigma: Standard deviation of the gaussian smoothing filter.

Examples