Development of a New Fractal Algorithm to Predict Quality Traits of MRI Loins
Traditionally, the quality traits of meat products have been estimated by means of physico-chemical methods. Computer vision algorithms on MRI have also been presented as an alternative to these destructive methods since MRI is non-destructive, non-ionizing and innocuous. The use of fractals to analyze MRI could be another possibility for this purpose. In this paper, a new fractal algorithm is developed, to obtain features from MRI based on fractal characteristics. This algorithm is called OPFTA (One Point Fractal Texture Algorithm). Three fractal algorithms were tested in this study: CFA (Classical fractal algorithm), FTA (Fractal texture algorithm) and OPFTA. The results obtained by means of these three fractal algorithms were correlated to the results obtained by means of physico-chemical methods. OPFTA and FTA achieved correlation coefficients higher than 0.75 and CFA reached low relationship for the quality parameters of loins. The best results were achieved for OPFTA as fractal algorithm (0.837 for lipid content, 0.909 for salt content and 0.911 for moisture). These high correlation coefficients confirm the new algorithm as an alternative to the classical computational approaches (texture algorithms) in order to compute the quality parameters of meat products in a non-destructive and efficient way.
KeywordsMRI Fractal Algorithms Quality traits Iberian loin
The authors wish to acknowledge the funding received from the FEDER-MICCIN Infrastructure Research Project (UNEX-10-1E-402), Junta de Extremadura economic support for research group (GRU15173 and GRU15113) and the COST association, Farm Animal Imaging action (FAIM) (COST-FA1102) (COST-STSM-FA1102-26642). We also wish to thank the Animal Source Foodstuffs Innovation Service (SiPA, Cáceres, Spain) from the University of Extremadura.
- 4.Cernadas, E., Antequera, T., Rodríguez, P.G., Durán, M.L., Gallardo, R., Villa, D.: Magnetic resonance imaging to classify loin from Iberian pig. In: Webb, G.A., Belton, P.S., Gil, A.M., Delgadillo, I. (Eds.) Magnetic Resonance Imaging in Food Science: A View to the Future. The Royal Society of Chemistry. Cambridge (2001)Google Scholar
- 6.Ávila, M.M., Durán, M.L., Antequera, T., Palacios, R., Luquero, M.: 3D reconstruction on mri to analyse marbling and fat level in iberian loin. In: Martí, J., Benedí, J.M., Mendonça, A.M., Serrat, J. (eds.) IbPRIA 2007. LNCS, vol. 4477, pp. 145–152. Springer, Heidelberg (2007). doi: 10.1007/978-3-540-72847-4_20 CrossRefGoogle Scholar
- 8.Pérez-Palacios, T., Caballero, D., Caro, A., Antequera, T.: Magnetic resonance imaging and computational texture features to predict moisture and lipid content of loins. In: IV Farm Animal Imaging Conference, Edinburgh, UK (2015)Google Scholar
- 12.Sun, J., Zhang, Y.B., Dahl, A.B., Conradsen, K., Juul Jensen, D.: Boundary fractal analysis of two cube-oriented grains in partly recrystallized copper. In: XVII International Conference on Texture of Materials, ICOTOM 2017, Dresden, Germany (2014)Google Scholar
- 18.Association of Official Analytical Chemist (AOAC): Official Methods of Analysis of AOAC International, 17th edn. AOAC International. Gaithersburg, Maryland, U.S.AGoogle Scholar
- 21.Caballero, D., Caro, A., Antequera, T., Pérez-Palacios, T.: Non destructive analysis of loin by magnetic resonance imaging and fractal. In: IX Sympossium of Mediterranean Pig, Portalegre, Portugal (2016)Google Scholar
- 22.Peckinpaugh, S.: An improved method for computing gray-level coocurrence matrix based texture measured. Comput. Vis. Graph. Image Process. 53, 574–580 (1991)Google Scholar
- 25.Kira, K., Rendell, L.A.: A practical approach to feature selection. In: IX International Conference on Machine Learning, Aberdeen, UK (1992)Google Scholar
- 30.Colton, T.: Statistics in Medicine. Little Brown and Co., New York (1974)Google Scholar