Total Variation Image Denoising
The Rudin-Osher-Fatemi total variation (TV) denoising technique poses the problem of denoising as a minimization problem:
Min int (f-u)^2 + lambda * int |grad u|
where f is the noisy image, lambda is a parameter, and u is the denoised image we seek. We may solve this minimization problem by solving the associated PDE:
u_t = (f-u) + (lambda/2) * curl u
u = tvdenoise(f,lambda,NumSteps) denoises the input image f over NumSteps number of iterations. The larger the parameter lambda, the stronger the denoising.
For a demo, run tvdenoise without any inputs.