ARTIFICIAL NEURAL NETWORK MODELING OF HYPERSPECTRAL RADIOMETRIC DATA FOR QUALITY CHANGES ASSOCIATED WITH AVOCADOS DURING STORAGE

Abstract

Artificial neural network (ANN) and hyperspectral techniques were used to model quality changes in avocados during storage at different temperatures. Avocados were coated using a pectin-based emulsion and stored at different temperatures (10, 15, 20C), along with uncoated control samples. At different time intervals during storage period, respiration rate, total color difference, texture and weight loss of samples were measured as conventional quality parameters. Hyperspectral imaging was used to evaluate spectral properties of avocados. Multilayer ANNs were used in two ways to develop models for predicting quality parameters during storage. In the first set, ANN models were developed based on principal components of hyperspectral data as well as storage temperature and time. The optimal configuration of neural network model was obtained by varying the different model parameters. Results indicated ANN models to be accurate and versatile and they predicted the quality changes in avocado fruits better than the conventional regression models; furthermore, the storage time–temperature-based ANN models were better than the hyperspectra-based ANN models. PRACTICAL APPLICATIONS The manuscript evaluates the use of artificial neural network (ANN) models to relate the postharvest quality of avocados to traditional process variables like storage time and temperature. To provide a new objective method, data were also gathered from nondestructive hyperspectral radiometric technique as opposed to simple linking of quality of avocados to storage variables. ANN models were trained using both sets of data and were compared. The study demonstrated the ANN models to be more accurate in predicting the quality changes in avocado fruits than the conventional regression models; and furthermore, the storage time–temperature-based ANN models were better than the hyperspectra-based ANN models. ANN techniques are currently being used in many food process-modeling applications. The study demonstrates the importance of new modeling techniques in predicting the quality of stored produce as well as the validity of well-recognized storage parameters like temperature and time to be more important in quality considerations.