| Grant number: | 13/22205-3 |
| Support type: | Scholarships abroad - Research Internship - Post-doctor |
| Effective date (Start): | April 01, 2014 |
| Effective date (End): | March 31, 2015 |
| Field of knowledge: | Physical Sciences and Mathematics - Computer Science |
| Principal Investigator: | Odemir Martinez Bruno |
| Grantee: | Joao Batista Florindo |
| Supervisor abroad: | Gabriel Landini |
| Home Institution: | Instituto de Física de São Carlos (IFSC). Universidade de São Paulo (USP). São Carlos , SP, Brazil |
| Local de pesquisa : | University of Birmingham, England |
| Associated to the scholarship: | 12/19143-3 - Fractal geometry and image analysis applied to vegetal biology, BP.PD |
Abstract This work proposes to investigate the identification of plant species by fractal-based imaging methods. Images of plant leaves will be obtained from three different modalities (i.e. scanned leaf surface, microscopy images of the stomata and histological cross-sections) and will be analyzed using innovative computerized methods developed for the analysis of the architecture of tumors and geometry of cancer cells as proposed by the supervisor at the University of Birmingham. These methods include asymptotic fractal dimensions, morphological characterization of cell neighborhoods and local connected fractal dimension. Other approaches studied by the post-doctoral researcher at his home institution will be applied, including fractal descriptors and multiscale fractal dimension. The rational of this study is that image analysis can provide a new perspective over taxonomical identification, integrating the analysis of the arrangement of micro and macro-structures on the leaves which express both evolutionary processes as well as the interrelationships between their biological structure and the environment. In this way, the proposed analyses can provide an unbiased and useful description of plant tissue samples to facilitate a precise and reliable categorization of species. This work will contribute to both image analysis community as well as botany and cancer researchers by developing inexpensive computational methodologies to extract quantitative data for sample classification purposes. (AU) | |