Quantifying the Morphology of Bread Crusts

A novel way to analyse and parametrise the structure of bread crust is presented. It uses digital photographs of slices of bread, from which grey value traces along curves parallel to the crust surface are extracted. From these traces intensity autocorrelation functions are calculated, and from these, distances that characterise the structure are determined. A computational technique to localise the crust surface and obtain its smoothed contour was developed. The method is shown to be able to quantify differences in average crust structure (between 0.2 and 5 mm from the crust surface) for some selected bread improvers. By comparing the crust at the top and the bottom of a roll, it was also demonstrated that the method is able to give spatially resolved structural information inside the crust. Separating systematic (e.g enzyme or improver type dependent) effects from ‘natural’ random variations between individual slices demands 10 to 40 independent images.