Image analysis

The analysis of a sample on the basis of its structure, as determined by non-destructive techniques such as microscopy or machine vision. Parameters of interest in the image can be both classified and quantified using the human eye or computer programs. In food plants it grades produce, measures fat-to-lean ratios in meat — for example, grading for speedy online determination of fat/lean ratio — and spots defects on sorting lines, bringing speed and objectivity to quality decisions once made by hand. Fibre-optic probes extend the approach to predicting textural defects such as excessive paleness or darkness in meat.