Fractal geometry may be used to represent geometric structures of great complexity, including phenomena in nature where the traditional geometries can not be used. The leaves and their structures, although are not perfect fractals, are one of these phenomena that can be described and characterized using fractal geometry. In this direction studies using fractals in plant identification by texture analysis has been carried out in recent years by the GCC group (Scientific Computing Group) of the IFSC-USP. These studies has been gaining attention in recent years with regard to its quality as a method of texture analysis. There is still, however, many ways to improve this computational method, being this one of main objectives of this project. In a complementary way it is intended to investigate the robustness of the descriptors with respect of fractal analysis species with leaf anatomical plasticity. The phenotypic plasticity in plants is defined as the change in the expression of the genotype by environmental influences, may present in both morphologically and physiologically. Such investigation have an particularly importantance since the automated systems of leaf identification must be able to deal with such morphological changes. Recent research results are presented in this project demonstrating the viability of these objectives.
News published in Agência FAPESP Newsletter about the scholarship:
AMANCIO, DIEGO RAPHAEL;
COMIN, CESAR HENRIQUE;
BRUNO, ODEMIR MARTINEZ;
RODRIGUES, FRANCISCO APARECIDO;
COSTA, LUCIANO DA FONTOURA.
A Systematic Comparison of Supervised Classifiers.
APR 24 2014.
Web of Science Citations: 67.
Please report errors in scientific publications list by writing to: