Signal and Image Processing

Elimination of shifts of photo image rows based on estimation in sliding window and regularized integration of second differences of the shifts

A. V. Masjukov, I. S. Murzov

Tver State University, Tver

Abstract: We consider an image reconstruction problem when the rows of a given image have unknown shifts. The proposed algorithm consists of evaluation and subsequent regularized integration of second differences of the shifts. The moving gate evaluation of the second differences is based on the assumption that level lines of an undamaged image are smoother. The regularized integration implies convolution with the proposed filter. The method is verified in a test with random shifts of the rows of a photo. The considered problem is a model problem for short-period static correction in seismic data processing. We propose a new algorithm for short-period static correction.

Keywords: digital signal processing, ill-posed problems, regularization.

UDC: 519.6+517.983.54+550.8

Received: 20.03.2009
Revised: 11.05.2009

Bibliographic databases:

•