High-speed assessment of bread quality through optimal non-contact methods

Being a bio-process, the quality of yeast-fermented bread is highly dependent on several factors such as temperature, static time, and humidity as well as on the type of flour. As such an active control/feedback strategy is often needed to monitor and optimize the baking process. Many studies have focused on post-baking the assessment of the breads using food’s color and other optical testing to implement a closed feedback control. Other studies have evaluated the height and slope of the bread to infer quality. Quality in food in general, and bread-making in particular is often judged subjectively by humans. As such machine learning methods are increasingly applied to quality inspection of food. There are several versatile machine learning tools such as k-nearest neighbor (KNN), neural networks, support vector machine (SVM), and artificial neural network (ANN).
Silver Hills Bakery is looking to significantly expand and scale up production. The current method of manual inspection and evaluation would not be sufficient for the expected increase in production. To this end, the project will develop an optimal non-contact methodology for estimating bread quality for controlling baking parameters

Faculty Supervisor:

Krishna Vijayaraghavan

Student:

Partner:

Silver Hills Bakery

Discipline:

Engineering

Sector:

Manufacturing

University:

Simon Fraser University

Program: