Abstract
In food industry, quality and safety parameters are direct related with consumers' health condition. There is a growing interest in developing non-invasive sensorial techniques that have the capability of predicting quality attributes in real-time operation. Among other detection methodologies, Fourier transform infrared (FTIR) spectroscopy has been widely used for rapid inspection of various food products. In this paper, an advanced clustering-based neurofuzzy identification model has been developed to detect meat spoilage microorganisms during aerobic storage at various temperatures, utilizing FTIR spectra. A clustering scheme has been utilized as an initial step for defining the fuzzy rules while an asymmetric Gaussian membership function has been used in the fuzzification part of the model. The proposed model not only classifies meat samples in their respective quality class (i.e. fresh, semi-fresh and spoiled), but also predicts their associated microbiological population directly from FTIR spectra. Results verified the superiority of the proposed scheme against the adaptive neuro-fuzzy inference system, multilayer perceptron and partial least squares in terms of prediction accuracy.