Estimation of Bread Surface Moisture: Rapid, Contactless Measurement

A major challenge in the industrial production of baked goods is monitoring moisture content, a critical measurement at all stages of production. Raw materials can contain water that may lead to clumping or degradation during storage, ultimately reducing shelf life and product value.

Moisture content can traditionally be determined using vacuum or conventional ovens. However, these methods only sample a small portion of the product population, and heating the samples can risk damaging the product. Hyperspectral imaging offers the possibility to estimate moisture content without the risk of degradation. Headwall’s industrial machine vision systems, in combination with PerClass Mira® analysis software, enable a contactless estimation of this crucial component within minutes, either inline or offline, using the PerClass Mira Scanning Stage.

Overview

In a recent study, water was applied to the surface of various slices of bread. Scans of the bread were taken using the PerClass Mira Software along with Headwall's MV.C NIR hyperspectral imaging sensor on the PerClass Mira Stage. This process was repeated over the course of an hour as the water evaporated. Based on these scans, a regression model was created to estimate the water content of the bread.

Estimate the water content of bread with Headwall's MV.C NIR hyperspectral imaging sensor.

Regression Modeling

Using hyperspectral imaging scans, a PerClass Mira classification model was trained to distinguish the background from the bread slice, segmenting the slice as an object. This object was then assigned a value from 0 to 60, indicating the number of minutes since the water was applied. The PerClass Mira software then created a regression model using the average spectra of the bread to predict the moisture content on the surface of other bread slices. In addition to estimating moisture content, the roundness and size of baked products can also be determined via the PerClass Mira software and exported downstream for inline action.

organic-brown-bread-slice-placed-wooden-board-rustic-style.jpg

Advantages of Hyperspectral Imaging

The results of this study demonstrate that PerClass Mira can determine surface moisture content without heating and potentially altering the product. Hyperspectral images can be scanned and analyzed in real time, providing quicker information about the product and enabling potential 100% product inspection to ensure quality. Headwall's MV.C NIR combined with PerClass Mira Software and Scanning Stage offers a tool for fast and contactless solutions tailored to your industrial baking needs.

Image by KamranAydinov on Freepik

More information?

Please contact our specialist in spectroscopy Kees van der Sar.

Bread surface measurements and/or bread core measurements;

Estimation of Bread Surface Moisture: Rapid, Contactless Measurement

Inventech is the specialist in spectrometry, also for bread moisture measurements.