Performance analysis of a maximum-likelihood speckle motion estimator

Presented herein is a performance analysis of a maximum likelihood estimator for calculating small speckle motions. Such estimators are important in a variety of speckle techniques used in non-destructive evaluation. The analysis characterizes the performance (bias and RMS deviation) of the estimator as a function of the signal-to-noise ratio. This SNR parameter is a convenient surrogate for decorrelation of sequential speckle patterns such as are seen in biological tissues. Although the particular estimator is predicated on speckle motions that are a small fraction of a pixel, accurate performance is demonstrated for instantaneous motions of up to +/-0.8 pixel/record. Beyond this point, more conventional approaches have been shown to perform well.