URBAN ENVIRONMENTAL QUALITY ASSESSMENT BY SHAPE AND SPECTRAL INDICES OF MULBERRY LEAVES

In this paper, an analysis of the potential use of the surface and geometric characteristics of mulberry leaves as parameters for environmental quality assessment is made. Methods have been used to reduce the amount of data of latent variables, linear and kernel variants of principal components. It has been found that a kernel variant of the principal components, combined with nonlinear separating functions of discriminant analysis and a method of support vector machines, are an appropriate methods for distinguishing the degree of air pollution from the mulberry leaf data. The results obtained could be used as preliminary baseline data for future evaluations and studies related to remote monitoring of urban air quality.