Abstract
The increasing consumption worldwide of tortilla chips make it relevant to design and optimize their industrial quality analysis. Surface, structural, and total oil content during frying of tortilla chips fried at 160, 175, and 190°C for different times were analyzed. The aim was to obtain a relationship between oil content and features from their digital images. The results showed a high linear correlation (R > 0.90) between oil content with image features at each frying temperature, indicating that trustable models can be developed, allowing the prediction of oil content of tortilla chips by using selected features extracted from their digital images, without the necessity of measuring them. Cross-validation technique demonstrated the repeatability of each model and their good performance (>90%).
ACKNOWLEDGMENTS
The authors would like to acknowledge Sr. Jon May from FHM Alimentos Ltd., for providing the masa to chips elaboration. This work was financially supported by Vicerrectoría de Investigación y Desarrollo of Universidad de Santiago de Chile (USACH), FONDECYT Project 1070031, LACCIR Virtual Institute (Project R0308LAC003), Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET) from Argentina, and PBCT-PSD-62 Project from CONICYT-Chile.
Notes
1Balu Toolbox can be downloaded from http://dmery.ing.puc.cl.