Prediction of Bread Crumb Density by Digital Image Analysis

Abstract

ABSTRACTThe cellular structure of bread crumb (crumb grain) is an important factor that contributes to the textural properties of fresh bread. The accuracy of a digital image analysis (DIA) system for crumb grain measurement was evaluated based on its capability to predict bread crumb density from directly computed structural parameters. Bread was prepared from representative flour samples of two different wheat classes, Canada Western Red Spring (CWRS) and Canada Prairie Spring (CPS). Dough mixing and proofing conditions were varied to manipulate loaf volume and crumb density. Sliced bread was subjected to DIA immediately after physical density measurement. Experiments were repeated for the same bread samples after drying to three different moisture contents. Five computed crumb grain parameters were assessed: crumb brightness, cell wall thickness (CWT), void fraction (VF), mean cell area, and crumb fineness (measured as number of cells/cm2). Crumb density ranged from 0.088 to 0.252 g/cm3 depending on proofing and mixing treatments, and was predominantly affected by the former. With increasing crumb density, bread crumb became brighter in appearance, mean cell size and CWT decreased, crumb fineness increased, and the VF decreased. Approximately 80% of the variation in fresh or dried crumb density could be predicted using a linear regression model with two variables, CWT and VF. Results indicated that DIA of directly computed crumb grain could accurately predict bread crumb density after images had been correctly classified into cells and background.