Abstract

IBANEZ-BARBOZA, María de Jesús et al. Effect of aluminum and pH on floral development and pigmentation of Hydrangea macrophylla through artificial neural networks. Rev. fitotec. mex [online]. 2023, vol.46, n.2, pp.187-194.  Epub Apr 02, 2024. ISSN 0187-7380.  https://doi.org/10.35196/rfm.2023.2.187.

In hydrangea (Hydrangea macrophylla), the color of inflorescences changes with development and effect of various factors. In this study, the effect of two stages of floral development (Fresh, Antique), four doses of Al2(SO4)3 and two pH (4.5, 6.0) on the color of the florets of two hydrangea cultivars: pink (TolR) and blue (TolA) was determined. A methodology with image analysis and artificial neural networks (ANN) was used to determine numerically color changes. A classifier created with artificial neural networks (ANN) was used; the RGB (red, green, blue) system was used as input variables and four color classes: blue, pink, brown and green were used as output. The variables calculated were surface by color class (SCC, %), frequency of the dominant color group (FCD, %), luminosity (CIE L*) and chroma (C*). The pink surface was the dominant class in both cultivars in the Fresh state with 8.5 % (pH 4.5) and 13 % (pH 6.0) and greater surface in TolR than in TolA, while the green (TolA) and brown (TolR) classes obtained greater surface (85%) in Antique state with no effect by dose of Al2(SO4)3 or pH. When applying 1 g L-1 of Al2(SO4)3 in TolA, the color group (RHS chart) with the highest CDF changed from purple-violet to red-purple under pH 6.0 in the Fresh state and from yellowish-green to greyish-brown with pH 4.5 in Antique state. The color variation was mainly due to the cultivar and floral development stage.

Keywords : Hydrangea macrophylla; artificial neural networks; color variables; image analysis.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )