Full text
| |
| Author(s): |
Núbia Rosa da Silva
Total Authors: 1
|
| Document type: | Doctoral Thesis |
| Press: | São Carlos. |
| Institution: | Universidade de São Paulo (USP). Instituto de Ciências Matemáticas e de Computação (ICMC/SB) |
| Defense date: | 2015-10-22 |
| Examining board members: |
Odemir Martinez Bruno;
Célia Aparecida Zorzo Barcelos;
Rosana Marta Kolb;
Hélio Pedrini;
Francisco Aparecido Rodrigues
|
| Advisor: | Odemir Martinez Bruno |
| Abstract | |
Pattern recognition of texture in images has been playing an important role in computer vision area. This is because the texture attribute can reveal intrinsic characteristics, making it possible to classify a set of similar images. Although the texture is studied for over half a century, there is still no consensus on its definition or even a method to extrac texture characteristics that is effective for all types of images. Moreover, literature methods globally analyze the texture patterns, whereas a picture displays a number of micropatterns which form a single homogenous global pattern of texture in the image. However, some types of image display heterogeneity in their composition, that is, the set of micropatterns in the image use to form more than one texture pattern within the same image. Such type of image led to the purpose of this research work. Regardless the feature extraction method used, considering the heterogeneity of the texture pattern in an image leads to better representation of its features. To further improve the analysis of heterogeneous texture patterns, three approaches are proposed: (i) lazy-patch, (ii) combination of models and (iii) texture modeling using cellular automata inspired by pitting corrosion. The results of applying these approaches in different sets of biology and nanotechnology images showed that the analysis of heterogeneous patterns results in better representation of images that have heterogeneous patterns of texture in your composition. (AU) | |
| FAPESP's process: | 11/21467-9 - Heterogeneous Pattern Recognition and its Applications in Biology and Nanotechnology. |
| Grantee: | Núbia Rosa da Silva |
| Support Opportunities: | Scholarships in Brazil - Doctorate |