This paper demonstrates the effectiveness and versatility of an iterative deconvolution algorithm in dealing with noise-rendered, truncated signals when signal averaging is not an option. An iterative deconvolution algorithm for system identification and signal restoration is presented, and its effectiveness and robustness are validated through the analysis of several artificially generated signals that are intended to mimic practically measured signals. Its application is intended for use in improving the quality of system identification by reducing the detrimental effect of information leakage caused by windowing. System identification was conducted for various scenarios, in which the input and output signals were rendered with noise and subjected to different truncation levels at the heads and/or tails. The ability of the algorithm to restore the truncated portion of signals is demonstrated. It is concluded that the algorithm has superior performance compared to currently available traditional approaches, such as the fast Fourier transform and autoregressive moving average methods.
Edge quality based iterative deconvolution algorithm for motion blurred gray scale images for geometrical measures