Iterative image restoration by a method of steepest descent

Iterative image restoration by a method of steepest descent has been effectively carried out by using the minicomputer-based digital image processing system. Features of the method are that convergence is promising for most practical cases, a convergence rate is realised which is faster than that obtained by the static solution, and use can be made of a simpler processing algorithm than in the approach which uses the conjugate gradient method. Usefulness of the method is shown through simulation experiments of restoration of 1-D and 2-D degraded images. And restoration of an actually degraded image can be made successfully by means of man-machine interactive techniques.