Abstract

A novel approach to joint blind estimation of additive noise variance and probability of impulsive noise occurrence in images is put forward. It is based on using a fractal Brownian model for description of real life image characteristics. As shown, this approach allows rather accurate estimation of mixed noise parameters even for images containing a large percentage of texture regions.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call