Uso del procesamiento digital de imágenes para la extracción de datos de medidas experimentales publicados en formato gráfico

Abstract

In this article, we propose a method as an alternative to obtain experimental measurement data, in the absence of laboratory equipment to perform tests, in a suitable format to perform mathematical operations in order to use them as information to validate: hypotheses, models constitutive and / or research theories focused on technological development. The proposed method uses as a main tool the image segmentation technique by region growth by pixel grouping and the normalization of the coordinates of the positions of the pixels extracted to the axis scale in the corresponding figure. The segmentation of the image separates the coordinates of the pixels that form the axes and the curves, the coordinates of the pixels of the curves are normalized to the scale of the axes. The method is tested with images of the result of experimental tests of stress-strain behavior recovered from [1]. The results of the data extraction are plotted and the averages of each curve extracted as well as the standard deviation are obtained. It is verified that the data obtained can be used to corroborate or support hypotheses in a wide range of investigations.