Pattern Recognition on Images Based on Complex Systems

• Grant number: 16/16060-0
• Support Opportunities: Regular Research Grants
• Start date: February 01, 2017
• End date: January 31, 2019
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computational Mathematics
• Principal Investigator: Joao Batista Florindo
• Grantee: Joao Batista Florindo
• Host Institution: Instituto de Matemática, Estatística e Computação Científica (IMECC). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil
• Associated researchers: Odemir Martinez Bruno ; Rosana Marta Kolb

Abstract

This project proposes the study and development of mathematical and computational tools for image analysis based on complex systems. More specifically, the researcher will focus on three approaches, namely, fractal geometry, complex networks and methods of Mathematical Physics. All these techniques share the ability of describing with great accuracy the irregularity or homogeneity of an object represented in an image viewed from different scale levels. This is crucial information even for our vision system, which frequently uses this information to distinguish objects and scenery around us. The same importance of this attribute can be observed in computer vision systems, particularly in the analysis of nature images, as demonstrated in several studies previously developed by the applicant. Finally, the techniques developed here will be applied to real-world problems involving biological images, especially in plant taxonomy and phenotyping. It is also worth noting that this project aims to create a symbiosis between application and theory and so both theories developed for general purpose applications will be tested on specific problems as well as such problems will also give rise to new theoretical lines that can meet their needs. This paper describes the theoretical lines and involved applications, as well as the expected scientific production and potential qualification of human resources. (AU)

Scientific publications

(14)

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

FLORINDO, JOAO BATISTA; VERIKAS, A; RADEVA, P; NIKOLAEV, D; ZHOU, J. Singular Spectrum Decomposition of Bouligand-Minkowski Fractal Descriptors: An Application to the Classification of Texture Images. TENTH INTERNATIONAL CONFERENCE ON MACHINE VISION (ICMV 2017), v. 10696, p. 8-pg., 2018-01-01. (16/16060-0)

SILVA, PEDRO M.; FLORINDO, JOAO B.. Using down-sampling for multiscale analysis of texture images. PATTERN RECOGNITION LETTERS, v. 125, p. 411-417, JUL 1 2019. (16/16060-0)

FLORINDO, JOAO B.; BRUNO, ODEMIR M.. Texture classification using non-Euclidean Minkowski dilation. PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS, v. 493, p. 189-202, MAR 1 2018. (16/16060-0, 12/19143-3, 14/08026-1)

FLORINDO, JOAO BATISTA; CASANOVA, DALCIMAR; BRUNO, ODEMIR MARTINEZ. A Gaussian pyramid approach to Bouligand-Minkowski fractal descriptors. INFORMATION SCIENCES, v. 459, p. 36-52, AUG 2018. (16/16060-0, 12/19143-3, 14/08026-1, 13/22205-3)

ESMI, E.; FLORINDO, J. B.; PEREZ, F.; BARBEITOS, M.; LIU, J; LU, J; XU, Y; MARTINEZ, L; KERRE, EE. An approach to recognize coral species on the coast of Brazil using image analysis and fuzzy associative memories based on equivalent measures. DATA SCIENCE AND KNOWLEDGE ENGINEERING FOR SENSING DECISION SUPPORT, v. 11, p. 8-pg., 2018-01-01. (16/26040-7, 16/16060-0)

FLORINDO, JOAO BATISTA; BRUNO, ODEMIR MARTINEZ. Fractal Descriptors of Texture Images Based on the Triangular Prism Dimension. Journal of Mathematical Imaging and Vision, v. 61, n. 1, p. 140-159, JAN 2019. (13/22205-3, 16/16060-0, 12/19143-3, 14/08026-1)

METZE, KONRADIN; ADAM, RANDALL; FLORINDO, JOAO BATISTA. The fractal dimension of chromatin - a potential molecular marker for carcinogenesis, tumor progression and prognosis. EXPERT REVIEW OF MOLECULAR DIAGNOSTICS, v. 19, n. 4, p. 299-312, APR 3 2019. (16/16060-0)

FLORINDO, JOAO B.; LEE, YOUNG-SUP; JUN, KYUNGKOO; JEON, GWANGGIL; ALBERTINI, MARCELO K.. VisGraphNet: A complex network interpretation of convolutional neural features. INFORMATION SCIENCES, v. 543, p. 296-308, JAN 8 2021. (16/16060-0)

TARASCHI, GIOVANNI; FLORINDO, JOAO B.. Computing fractal descriptors of texture images using sliding boxes: An application to the identification of Brazilian plant species. PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS, v. 545, MAY 1 2020. (16/16060-0)

SILVA, PEDRO M.; FLORINDO, JOAO B.. A statistical descriptor for texture images based on the box counting fractal dimension. PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS, v. 528, AUG 15 2019. (16/16060-0)

FLORINDO, JOAO BATISTA; BRUNO, ODEMIR MARTINEZ. Discrete Schroedinger transform for texture recognition. INFORMATION SCIENCES, v. 415, p. 142-155, NOV 2017. (16/16060-0, 12/19143-3, 14/08026-1)

SILVA, PEDRO M.; FLORINDO, JOAO B.. Fractal measures of image local features: an application to texture recognition. MULTIMEDIA TOOLS AND APPLICATIONS, v. 80, n. 9, p. 14213-14229, APR 2021. (16/16060-0)

FLORINDO, JOAO B.; LAUREANO, ESTEVAO ESMI. BoFF: A bag of fuzzy deep features for texture recognition. EXPERT SYSTEMS WITH APPLICATIONS, v. 219, p. 12-pg., 2023-02-03. (16/16060-0)

VIEIRA, JARDEL; ABREU, EDUARDO; FLORINDO, JOAO B.. Texture image classification based on a pseudo-parabolic diffusion model. MULTIMEDIA TOOLS AND APPLICATIONS, v. 82, n. 3, p. 24-pg., 2022-07-11. (16/16060-0, 19/20991-8)