Cell phone-based two-dimensional spectral analysis for banana ripeness estimation

Abstract

It is known that when green fruits such as bananas and mangoes grow from the immature to the ripe stages, their color changes from green to yellow. However, once the green fruits are yellow, it is difficult to classify them into the ripe and the overripe stages. Here, we propose and experimentally demonstrate for the first time a two-dimensional banana ripeness level sensor by using a cell phone. Our key approach relies on simultaneous analysis of two broad-spectral images of the banana under white light and ultraviolet illuminations. Particularly, our selected color ratios from these two broad-spectral images are used for spatially and specifically classifying the whole banana into immature, ripe, and overripe zones. Experimental proof of concept using a smart mobile phone is highlighted, featuring fast response, portability, and ease of implementation. © 2012 Elsevier B.V.