Determinação do teor alcoólico de cachaças por Imagens Digitais

Abstract

This work proposes the use of digital images obtained by scanner to determine the alcohol content in cachaça. Obtaining the image of an object by scanner generally uses a bar of radiation reaching the image that reflects radiation to the sensor allowing the reconstruction and storage of digital images. Each image generated in accordance with the red-green-blue system (RGB) provides an array of values of RGB components. The standard of RGB vectors was adopted as analytical response to the construction of titration curves based on digital images. To evaluate the results with those obtained by Gay-Lussac Alcoholmeter with digital images, a paired t test was applied to the dataset to a confidence level of 95%. It was found that there was no significant difference between the two methodologies for the confidence level seen (p <0.05). The relative standard deviations of the results obtained by the proposed method were less than 3.0%.