TY - GEN
T1 - Assessing the quality level of corn tortillas with inductive characterization and digital image analysis
AU - Moreno-Armendáriz, Marco A.
AU - Godoy-Calderon, Salvador
AU - Calvo, Hiram
AU - Rojas-Padilla, Oscar M.
N1 - Funding Information:
The authors wish to thank the support of the Mexican Government (CONACYT, SNI, SIP-IPN, COFAA-IPN, and PIFI-IPN).
PY - 2013
Y1 - 2013
N2 - Characterization and classification of corn tortillas turns out to be an extremely delicate and difficult process when dealing with regulations for import/export and production process certification. In this paper we present a method for non-invasive feature extraction, based on digital imaging and a series of procedures to characterize different qualities of corn tortillas for their later classification. The novelty in this whole method lies in the extremely reduced set of features required for the characterization with only geometrical and color features. Nonetheless, this set of features can assess diverse quality elements like the homogeneity of the baking process and others alike. Experimental results on a sample batch of 600 tortillas show the presented method to be around 95% effective.
AB - Characterization and classification of corn tortillas turns out to be an extremely delicate and difficult process when dealing with regulations for import/export and production process certification. In this paper we present a method for non-invasive feature extraction, based on digital imaging and a series of procedures to characterize different qualities of corn tortillas for their later classification. The novelty in this whole method lies in the extremely reduced set of features required for the characterization with only geometrical and color features. Nonetheless, this set of features can assess diverse quality elements like the homogeneity of the baking process and others alike. Experimental results on a sample batch of 600 tortillas show the presented method to be around 95% effective.
KW - Inductive characterization
KW - corn tortilla
KW - digital image analysis
UR - http://www.scopus.com/inward/record.url?scp=84888249676&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-38989-4_5
DO - 10.1007/978-3-642-38989-4_5
M3 - Contribución a la conferencia
SN - 9783642389887
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 40
EP - 53
BT - Pattern Recognition - 5th Mexican Conference, MCPR 2013, Proceedings
T2 - 5th Mexican Conference on Pattern Recognition, MCPR 2013
Y2 - 26 June 2013 through 29 June 2013
ER -