Gilberto Camara

Parameter selection for region‐growing image segmentation algorithms using spatial autocorrelation

Region‐growing segmentation algorithms are useful for remote sensing image segmentation. These algorithms need the user to supply control parameters, which control the quality of the resulting segmentation. An objective function is proposed for selecting suitable parameters for region‐growing algorithms to ensure best quality results. It considers that a segmentation has two desirable properties: each of the resulting segments should be internally homogeneous and should be distinguishable from its neighbourhood. The measure combines a spatial autocorrelation indicator that detects separability between regions and a variance indicator that expresses the overall homogeneity of the regions.

DMSP/OLS night¿time light imagery for urban population estimates in the Brazilian Amazon

This article analyses DMSP/OLS night‐time imagery as an information source to detect human settlements and to estimate the urban population in the Amazon region. DMSP/OLS single orbits were used to generate a DMSP stable light mosaic for 2002, in which most of the urban settlements with a population higher than 5000 inhabitants were precisely identified. DMSP/OLS night‐time mosaic images from 1995, 1999 and 2002 were integrated with the IBGE census data and the correlation between DMSP/OLS night‐time light area and the urban population was compared. Coefficients of determination higher than 0.8 were obtained from the linear regression between DMSP/OLS night‐time lights and urban population census data. Although the fieldwork showed that DMSP image data could only record urbanized settlements with more than 2.5 km2 of well‐lit surface areas, the initial and final extension of the night‐time light foci were actually precisely registered. Therefore, this paper identifies the potential of DMSP night‐time light images for estimating urban population as well as the technical limitations of using such images as a means to monitor urban population dynamics annually in a region where data are scarce and the demographic dynamics are unique, as in the Brazilian Amazon.

Tabu Search Heuristic for Point-Feature Cartographic Label Placement

The generation of better label placement configurations in maps is a problem that comes up in automated cartographic production. The objective of a good label placement is to display the geographic position of the features with their corresponding label in a clear and harmonious fashion, following accepted cartographic conventions. In this work, we have approached this problem from a combinatorial optimization point of view, and our research consisted of the evaluation of the tabu search (TS) heuristic applied to cartographic label placement. When compared, in real and random test cases, with techniques such as simulated annealing and genetic algorithm (GA), TS has proven to be an efficient choice, with the best performance in quality. We concluded that TS is a recommended method to solve cartographic label placement problem of point features, due to its simplicity, practicality, efficiency and good performance along with its ability to generate quality solutions in acceptable computational time.

Remote-sensing image mining: detecting agents of land-use change in tropical forest areas

Land remote‐sensing images are the primary means of assessing land change. There have been major land changes in the planet in the last decades, especially in tropical forest areas. Identifying the agents of deforestation is important for establishing public policies that can help preserve the environment. This paper proposes a method for detecting the agents of land change in remote‐sensing image databases. We associate each land‐change pattern, detected in a remote‐sensing image, to one of the agents of change. The proposed method uses a decision‐tree classifier to describe shapes found in land‐use maps extracted from remote‐sensing images and then associates these shape descriptions to the different types of social agents involved in land‐use change. We support our proposal with two case studies for detecting land‐change agents in Amazonia, using the remote‐sensing image database of the Brazilian National Institute for Space Research (INPE).

A Framework for Measuring the Interoperability of Geo-Ontologies

Interoperability is a crucial problem for geographic information systems. The transfer of data and models between different systems requires the ability to set up a correspondence between concepts in one system to concepts in the other. Concept matching is helped by ontologies. However, the challenge of making ontologies themselves interoperable continues. In other words, given two geo-ontologies, the basic question is: to which degree are these two geo-ontologies interoperable? In this paper, we consider that a geo-ontology describes things that can be assigned to locations on the surface of the Earth and relations between these things. A geo-ontology has concepts that correspond to physical and social phenomena in the real world. We suggest a classification of these concepts based on their use for describing geo-objects. We present a basic set of concepts for a geographical ontology, based on descriptions of the physical world and of the social reality. We also present a framework for measuring the degree of interoperability between geo-ontologies. We consider that this problem is a special case of Bernsteins model management algebra for metadata descriptions. We propose to use a matching operator for measuring interoperability between ontologies. The proposed framework provides a first basis for computational tools that allow a more precise response to problem of ontology interoperability.

Mining patterns of change in remote sensing image databases

Remote sensing image databases are the fastest growing archives of spatial information. However, we still have a limited capacity for extracting information from large remote sensing image databases. There are currently very few techniques for image data mining and information extraction in large image data sets, and thus we are failing to exploit our large remote sensing data archives. This paper proposes a methodology to provide guidance for mining remote sensing image databases. The basic idea is to use domain concepts to build generic description of patterns in remote sensing images, and then use structural approaches to identify such patterns in images. We illustrate our proposal with a case study for detecting land use patterns in Amazonia from INPEs remote sensing image database.

Public Commons of Geographic Data: Research and Development Challenges

Across the globe individuals and organizations are creating geographic data work products with little ability to efficiently or effectively make known and share those digital products with others. This article outlines a conceptual model and the accompanying research challenges for providing easy legal and technological mechanisms by which any creator might affirmatively and permanently mark and make accessible a geographic dataset such that the world knows where the dataset came from and that the data is available for use without the law assuming that the user must first acquire permission.

What's in an Image?

This paper discusses the ontological status of remote sensing images, from a GIScience perspective. We argue that images have a dual nature--they are fields at the measurement level and fiat objects at the classification level--and that images have an ontological description of their own, distinct and independent from the domain ontology a domain scientist uses. This paper proposes a multi-level ontology for images, combining both field and object approaches and distinguishing between image and user ontologies. The framework developed contributes to the design of a new generation of integrated GISs, since two key benefits are achieved: (1) the support for multiple perspectives for the same image and (2) an emphasis on using images for the detection of spatial-temporal configurations of geographic phenomena.

Using fuzzy cognitive maps to describe current system dynamics and develop land cover scenarios: a case study in the Brazilian Amazon

In this study we developed a methodology to identify and quantify the relationships among determinants of land cover change using a regional case study in the Brazilian Amazon. The method is based on the application of fuzzy cognitive maps (FCMs), a semi-quantitative tool that provides a structured assessment of key feedbacks in scenario analysis. Novel to the application of FCMs is the use of spatial data-sets as the main input to build a cognitive map. Identification of interactions between land cover determinants and strengths is based on an empirical analysis of spatially explicit data and literature review. Expert knowledge is adopted to identify the strengths and weaknesses of the method. Potential pitfalls, such as spatial autocorrelation and scale issues, identified are intrinsic to the empirical data analysis. The outputs of the resulting FCMs are compared to the outputs of spatially explicit models under similar scenarios of change. The proposed method is said to be robust and reproducible when compared with participatory approaches, and it can endorse the consistency between demand and allocation in scenario analysis to be used in spatially explicit models.

An extensible toolbox for modeling nature-society interactions

Modeling interactions between social and natural systems is a hard task. It involves collecting data, building up a conceptual approach, implementing, calibrating, simulating, validating, and possibly repeating these steps again and again. There are different conceptual approaches proposed in the literature to tackle this problem. However, for complex problems it is better to combine different approaches, giving rise to a need for flexible and extensible frameworks for modeling nature-society interactions. In this paper we present TerraME, an open source toolbox that supports multi-paradigm and multi-scale modeling of coupled human-environmental systems. It enables models that combine agent-based, cellular automata, system dynamics, and discrete event simulation paradigms. TerraME has a GIS interface for managing real-world geospatial data and uses Lua, an expressive scripting language. TerraME is a toolbox for modeling and simulation of nature-society interactions.Novel abstractions and services support multiscale spatiotemporal modeling.It allows the combined use of several paradigms for model implementation.It provides an extensible high-level modeling language.GIS integration supports real-world case studies.