Measure Correlations

Only an imaging system can look at a sample particle by particle, and record data on each particle. This is its great advantage, and measure correlations are one way of exploiting that advantage to get maximum information from the data.

The simple example below shows a powder in which Bounding Rectangle Width and Length generally correlate, but there is some variation in aspect ratio (first plot).

In the second plot, the larger particles (above about 20 microns) had Smoothness mostly between 0.5 and 0.75, and the smaller particles show a larger spread in the Smoothness values.

The sample represented below shows a bimodal number distribution. Does this mean it contains two different kinds of particles? If we correlate ECA diameter to Ellipse aspect ratio, we see that objects in the higher peak are rounder – aspect ratios below 2.0, while some of the particles in the lower peak have higher aspect ratios–they are more elongated.