Windowing techniques for image restoration

Abstract

The large errors observed in inverse filter or Wiener filter restorations of images are mainly due to the fact that only a truncated region of image data is available for processing. In the earlier literature, it was suggested that the well-known time-series windows may be generalized for treating these errors. This paper examines the windowing technique for the restoration of general blurred images. Mathematical expressions for the restoration errors that arise from truncated data are derived. Optimal windows for image restoration are then designed on the basis of these expressions. With these optimal windows, near-perfect restorations can be obtained if the images vary gradually in intensity near their borders. Restorations using the optimal windows and some well-known time-series analysis windows are presented for comparison of their performance. © 1991.