SCDMImage: um aplicativo para detecção de características morfológicas em produtos agrícolas

Abstract

<p><span class="fontstyle0">With the evolution of cultivation techniques, the world production of fruits and vegetables has been presenting a continuous growth. Despite the increase in production, a significant portion of the harvested products is mainly lost in the post-harvest due to several factors, such as inapropriate handling, transportation and conservation, lack of selection and classification, poorly sized packaging, among others. In general, the quality of fruits and vegetables is usually evaluated by the following parameters: size, shape, mass, color, firmness, ripeness, absence of bruises and pathologies.The objective of this work was to develop an application in Matlab to detect morphological characteristics (width, length, area, perimeter) of Mangifera indica L., ‘Tommy Atkins’ mangos from images, and using the same images, estimate its weigth. The images of the mangoes were acquired with a digital camera attached to an illuminated box and later processed in an application developed in Matlab. The images showed the morphological characteristics of the samples were obtained and an equation was determined to estimate their mass. The determination of the mass from the images had a 96% correlation with the actual mass measured with a scale and it was possible to obtain a reliable estimate of the masses of the mangoes from their images.</span></p>