| Grant number: | 18/01033-3 |
| Support type: | Regular Research Grants |
| Duration: | April 01, 2018 - September 30, 2020 |
| Field of knowledge: | Physical Sciences and Mathematics - Geosciences |
| Principal Investigator: | Rogério Galante Negri |
| Grantee: | Rogério Galante Negri |
| Home Institution: | Instituto de Ciência e Tecnologia (ICT). Universidade Estadual Paulista (UNESP). Campus de São José dos Campos. São José dos Campos , SP, Brazil |
| Assoc. researchers: | Alejandro César Frery Orgambide ; Tatiana Sussel Gonçalves Mendes |
Abstract
Remote Sensing has become an important tool for environmental processes analysis. Among several applications, change detection using Remote Sensing imagery is a topic of great interest. Perform this application in a temporal and accurate way is extremely important to understand the relations between anthropic and natural phenomena, allowing then better decision making. Use image classification as an intermediate step for change detection has been shown as an appropriate procedure. However, this approach depends on the accuracy of the classification results, which consequently motivates the development of more accurate classification methods. Based on the concept of divergence, stochastic distances have received great attention in recent years. In the light of this discussion, this project proposes the development of new change detection methods with basis on concepts derived from classification and stochastic distance able to produce robust and competitive results in comparison to conventional methodologies. Analysis with simulated data and practical applications will be carried in order to validate proposed methods. (AU)