Abstract: This study aimed to rapidly and quantitatively characterize subtle changes in the water distribution uniformity of noodle dough sheet by hyperspectral imaging technology combined with distributional homogeneity index (DHI) method. Firstly, noodle dough sheets with different water content gradients were set up, and different spectral pretreatment methods and pattern recognition methods were used to establish water prediction models. Then, visualization of the water content distribution in dough sheet was achieved using hyperspectral imaging technology. Finally, DHI was used to quantitatively characterize the water distribution uniformity in dough sheet at different mixing times. The results showed that the model constructed by least-squares support-vector machine (LS-SVM) based on spectra pre-processed by standard normal variate (SNV) had the best prediction performance. The water distribution was the most uniform when the dough mixing time was 15 min. This study verified the feasibility of using hyperspectral imaging technology combined with DHI for measuring the water distribution homogeneity of dough sheet, which would provide a theoretical guidance for the automatic monitoring of the noodle dough mixing process.

Key words: hyperspectral; dough sheet; water distribution homogeneity; distributional homogeneity index; automatics monitoring