Advanced search
Start date
Betweenand

Anomaly detection, analysis and localization: a case study on digital static images from remote sensing applied to cartography

Grant number: 16/24185-8
Support type:Scholarships abroad - Research
Effective date (Start): February 01, 2018
Effective date (End): July 31, 2018
Field of knowledge:Interdisciplinary Subjects
Principal Investigator:Maurício Araújo Dias
Grantee:Maurício Araújo Dias
Host: Josef Kittler
Home Institution: Faculdade de Ciências e Tecnologia (FCT). Universidade Estadual Paulista (UNESP). Campus de Presidente Prudente. Presidente Prudente , SP, Brazil
Local de pesquisa : University of Surrey, England  

Abstract

For Cartography, which is a Geoscience related to mappings, an anomaly is a non-conforming pattern, such as an unknown or unexpected structure, for example, present in images captured by remote sensing, that needs to receive some kind of treatment to be solved in the images of this area of study. The vast quantity of these images, in large sizes, which are available for research, represent a challenge when using Digital Image Processing (DIP) as the only tool set to help treating anomalies, since the amount of data to be processed for these images requires many computationally expensive DIP operations. Moreover, researchers often treat anomalies using complex DIP algorithms, which do not always present accurate results. Without using any other tool to help treating anomalies, the image processing takes a long time and computational effort, besides being hard to achieve accuracy. Related to these drawbacks, the scientific literature presents many publications describing methods or algorithms based on the combination of Pattern Recognition (PR) tools, such as anomaly detection and analysis or anomaly detection and localization. Each pair of these tools can solve subsets of these drawbacks, but not all of them. Therefore this project aims at developing an anomaly detection, analysis and localization algorithm to automatically detect occurrences, identify types and locate spatial locations of anomalies, as pre-processing and cascade tasks before applying DIP operations to treat anomalies in digital static images from remote sensing applied to cartography. The proposed algorithm deals with all the aforementioned drawbacks. Therefore, it is very important take the wide experience and excellence of the Centre of Vision, Speech and Signal Processing, at the University of Surrey, as basis for the development of this study in order to create an algorithm that will help DIP operations and algorithms to treat anomalies with less time and computational effort and more accuracy.