An Time Domain Analysis Approach to Measurement of Three Dimensional Shape Distortion

Abstract

Autocorrelation function and cross-correlation function can be used to illustrate the dependency relation of random processes. The degree of mini-transmutation is often too small to detect. A new approach via utilization of correlation coefficient between two different signals in time domain was proposed to measure the tiny distortion of three dimensional objects. The correlation analysis and deduction on transformation from 3-D surface distribution to 1-D time series signal were made rationally. Numeric simulation experiments of discrete digital signal generated from triangular function composite generator were done and six groups of correlation coefficient curves were given deliberately with different signal length. The correlation and similarity degree among different similar waveforms were analyzed in detail. It shows that the approach is valid to detect the degree of distortion between the original standard one and the practical one.