Hybrid Optimum Signal Processing for a Strain Sensor

Recently there has been a shift in material sciences towards the use of non contacting laser optical methods to characterize material behaviour. This development has been driven by the need to test specimens at relatively high temperatures. We present a hybrid system for tracking laser speckles consisting of a Fourier-optical preprocessor followed by a digital signal processing unit that is able to determine engineering strain within specimen at viewing distances of several hundred millimeters. We will show in this paper that advanced digital signal processing using a polarity coincidence correlation method allows for an increase of the processing speed without undue degrading of system performance. We will further show that an optimum analog to digital conversion characteristic exists which if applied results in displacement estimates with almost ideal variances.