Potts Model - The Potts Model in Signal and Image Processing

The Potts Model in Signal and Image Processing

The Potts model has applications in signal reconstruction. Assume that we are given noisy observation of a piecewise constant signal To recover from the noisy observation vector, on seeks a minimizer of the corresponding inverse problem, the -Potts functional which is defined by

$P_\gamma(u) = \gamma \| \nabla u \|_0 + \| u-f\|_p^p = \gamma \# \{ i : u_i \neq u_{i+1} \} + \sum_{i=1}^n |u_i - f_i|^p$

The jump penalty forces piecewise constant solutions and the data term couples the minimizing candidate to the data The parameter controls the tradeoff between regularity and data fidelity. There are fast algorithms for the exact minimization of the L^1 and the L^2-Potts functional (Friedrich, Kempe, Liebscher, Winkler, 2008).

In image processing, the Potts functional is related to the segmentation problem. However, in two dimensions the problem is NP-hard (Boykov, Veksler, Zabih, 2001).

Famous quotes containing the words model, signal and/or image:

“If the man who paints only the tree, or flower, or other surface he sees before him were an artist, the king of artists would be the photographer. It is for the artist to do something beyond this: in portrait painting to put on canvas something more than the face the model wears for that one day; to paint the man, in short, as well as his features.”
—James Mcneill Whistler (1834–1903)

“By day thy warning ringing bell to sound its notes,
By night thy silent signal lamps to swing.”
—Walt Whitman (1819–1892)

“An image of its state;
The wings half spread for flight,
The breast thrust out in pride
Whether to play, or to ride
Those winds that clamour of approaching night.”
—William Butler Yeats (1865–1939)