Services on Demand
Journal
Article
Indicators
- Cited by SciELO
Related links
- Similars in SciELO
Share
RIA. Revista de investigaciones agropecuarias
On-line version ISSN 1669-2314
Abstract
CLEVA, M.S; SAMPALLO, G.M; GONZALEZ THOMAS, A.O and ACOSTA, C.A. Método para la determinación del volumen de una muestra de granos de arroz mediante el procesamiento digital de imágenes. RIA. Rev. investig. agropecu. [online]. 2013, vol.39, n.2, pp.185-190. ISSN 1669-2314.
This paper presents a method to determine the volume of a sample of rice kernels based on digital image processing, characterized by its simplicity and minimum requirements. Image registration of the sample is done with a flatbed scanner. We propose a model in which each kernel of the sample is treated as an ellipsoid with principal axis L (long), A (width) and E (thickness). The values L and A are determined from the ellipse that best fits the projected area of each kernel in the image. It is assumed that the width is proportional to the thickness of the grain (E = kA). The proportionality constant k is determined by equating the expression of sample volume of n kernels in terms of L, A and k with the actual volume experimentally obtained by the method of displacement of toluene. Determinations were performed on samples of four varieties with different compositions of entire and broken kernels. The k values obtained for the different varieties and compositions were very similar: 0,63 ± 0,01. With the value of k for each variety and averaging of all varieties, the volume was determined and are contrasted with that obtained by the method of displacement of toluene, being the difference, in general, less than 3%.
Keywords : Cereals; Morphology; Computer vision.