Convergence of recursive image restoration algorithms

Recursive image restoration is a powerful method for image restoration. Two-dimensional recursive filters usually require less computational time and storage requirements than nonrecursive algorithms and can handle space-variant blur easily. The convergent properties of a recursive algorithm, based on constrained deconvolution, are examined. Two versions of the algorithm are compared, and it is shown that a computationally fast version would be suitable for efficient practical image restoration.<
>