Prediction of banana quality attributes and ripeness classification using artificial neural network

S.E. Adebayo, N. Hashim, K. Abdan, M. Hanafi, M. Zude-Sasse
Laser light backscattering imaging (LLBI) with five laser diodes emitting at wavelengths 532, 660, 785, 830, and 1060 nm were employed for predicting quality attributes of banana fruit. The predicted attributes were chlorophyll, elasticity and soluble solids content (SSC). Classifications were done on six ripening stages from ripening stages 2 to 7. The prediction and classification models were built using an artificial neural network (ANN). The results indicated that measurement at 532 nm gave the highest correlation coefficient with 0.949 for chlorophyll prediction, while correlation coefficients of 0.862, 0.867 were the highest obtained for elastic modulus, SSC at 785 and 830 nm, respectively. 95.5% correct classification accuracy was obtained at 830 nm by use of the ANN classification model. The results showed that LLBI with an ANN can be used for non-destructive estimation of banana quality attributes and the subsequent ripeness classification.
Adebayo, S.E., Hashim, N., Abdan, K., Hanafi, M. and Zude-Sasse, M. 2017. Prediction of banana quality attributes and ripeness classification using artificial neural network. Acta Hort. (ISHS) 1152:335-344
http://www.actahort.org/books/1152/1152_45.htm
laser diodes, chlorophyll, soluble solid content, quality attributes, neural network
English

Acta Horticulturae