Digital image analysis to assess the texture of bread products

• MATLAB-based Bread Texture Analyser quantifies bread texture non-destructively. • Digital ‘texture-fingerprints’ objectively assess bread quality and uniformity. • Applied to five breads, capturing unique porosity, density, and cell distribution. • Digital analysis validated against sensory evaluation for perceived bread quality. • The quick, scalable, and cost-efficient method to monitor bread production. Texture is a key characteristic of bread, offering insights into its quality, ingredients, and baking methods. Digital image texture analysis involves computer analysis of the bread image taken with a digital camera or smartphone. We developed the Bread Texture Analyser, an application that facilitates bread texture digital image analysis and performs a fast and non-destructive assessment. The software generated a list of parameters unique to each bread texture, or ‘digital texture-fingerprints’. Additionally, the software produced an image where a colourmap technique was applied to visualise the connected regions by area size. The ‘fingerprint’ serves as a benchmark for monitoring baking process stability and product quality in production. We undertook research to examine the opportunities of using ‘digital texture-fingerprints’ and in the first case study, we applied image texture analysis to analyse the ‘fingerprints’ of the five distinctive types of bread products. In the second case study, we applied the digital image texture analysis to evaluate the artisan baguettes baked from the five sample flours. We used the digital texture-fingerprints as one of the tools to calibrate the baking performance of these flours. We found that digital image analysis can efficiently inspect bread texture, requiring much less effort and minimal running cost. We confirmed that the methodology works equally well on a diverse range of breads, where recipe-dependent distinctive textures are acquired as a result of using various ingredients and baking processes. The outcome of the study confirmed that digital image analysis is an acceptable method to inspect the bread texture. Furthermore, the methodology can be scaled up and deployed within the industry, as well as being applied to products other than breads.