ISPRS Int. J. Geo-Inf.2014, 3(4), 1491-1511; doi:10.3390/ijgi3041491 - published 22 December 2014 Show/Hide Abstract
Abstract: With the ever increasing volume of remote sensing imagery collected by satellite constellations and aerial platforms, the use of automated techniques for change detection has grown in importance, such that changes in features can be quickly identified. However, the amount of data collected surpasses the capacity of imagery analysts. In order to improve the effectiveness and efficiency of imagery analysts performing data maintenance activities, we propose a method to predict relevant changes in high resolution satellite imagery based on human annotations on selected regions of an image. We study a variety of classifiers in order to determine which is most accurate. Additionally, we experiment with a variety of ways in which a diverse set of training data can be constructed to improve the quality of predictions. The proposed method aids in the analysis of change detection results by using various classifiers to develop a relevant change model that can be used to predict the likelihood of other analyzed areas containing a relevant change or not. These predictions of relevant change are useful to analysts, because they speed the interrogation of automated change detection results by leveraging their observations of areas already analyzed. A comparison of four classifiers shows that the random forest technique slightly outperforms other approaches.
ISPRS Int. J. Geo-Inf.2014, 3(4), 1445-1490; doi:10.3390/ijgi3041445 - published 19 December 2014 Show/Hide Abstract
Abstract: Modern 3D geovisualization systems (3DGeoVSs) are complex and evolving systems that are required to be adaptable and leverage distributed resources, including massive geodata. This article focuses on 3DGeoVSs built based on the principles of service-oriented architectures, standards and image-based representations (SSI) to address practically relevant challenges and potentials. Such systems facilitate resource sharing and agile and efficient system construction and change in an interoperable manner, while exploiting images as efficient, decoupled and interoperable representations. The software architecture of a 3DGeoVS and its underlying visualization model have strong effects on the system’s quality attributes and support various system life cycle activities. This article contributes a software reference architecture (SRA) for 3DGeoVSs based on SSI that can be used to design, describe and analyze concrete software architectures with the intended primary benefit of an increase in effectiveness and efficiency in such activities. The SRA integrates existing, proven technology and novel contributions in a unique manner. As the foundation for the SRA, we propose the generalized visualization pipeline model that generalizes and overcomes expressiveness limitations of the prevalent visualization pipeline model. To facilitate exploiting image-based representations (IReps), the SRA integrates approaches for the representation, provisioning and styling of and interaction with IReps. Five applications of the SRA provide proofs of concept for the general applicability and utility of the SRA. A qualitative evaluation indicates the overall suitability of the SRA, its applications and the general approach of building 3DGeoVSs based on SSI.
ISPRS Int. J. Geo-Inf.2014, 3(4), 1412-1444; doi:10.3390/ijgi3041412 - published 18 December 2014 Show/Hide Abstract
Abstract: This article is based on a study of the morphological changes of Dhaka City, the capital of Bangladesh. The main objective of the research is to study the transformation of urban morphology in Dhaka City from 1947 to 2007. Three sample wards (18, 19 and 72) of Dhaka City Corporation are strategically selected as the study areas. Ward 72 has an indigenous type of organic settlement, whereas ward 19 is a planned area, and ward 18 represents a mixed (both planned and informal) type of settlement. In this research, the transformation of urban settlement pattern is examined through space syntax. The results show that the organic settlements (ward 72) are highly integrated both in terms of the local and global syntactic measures (lowest standard deviation for local and global integration, with the highest intelligibility values), and are more connectivity. The scenario is opposite in the case of planned settlements. The characteristics of mixed areas (ward 18) lie in between the organic and planned settlements. Therefore, in summary, it can be stated that the integration, connectivity and intelligibility measures of Dhaka City are found to be high, medium and low for the indigenous, mixed and planned settlement types; respectively.
ISPRS Int. J. Geo-Inf.2014, 3(4), 1387-1411; doi:10.3390/ijgi3041387 - published 11 December 2014 Show/Hide Abstract
Abstract: Logistic regression is a classical linear model for logit-transformed conditional probabilities of a binary target variable. It recovers the true conditional probabilities if the joint distribution of predictors and the target is of log-linear form. Weights-of-evidence is an ordinary logistic regression with parameters equal to the differences of the weights of evidence if all predictor variables are discrete and conditionally independent given the target variable. The hypothesis of conditional independence can be tested in terms of log-linear models. If the assumption of conditional independence is violated, the application of weights-of-evidence does not only corrupt the predicted conditional probabilities, but also their rank transform. Logistic regression models, including the interaction terms, can account for the lack of conditional independence, appropriate interaction terms compensate exactly for violations of conditional independence. Multilayer artificial neural nets may be seen as nested regression-like models, with some sigmoidal activation function. Most often, the logistic function is used as the activation function. If the net topology, i.e., its control, is sufficiently versatile to mimic interaction terms, artificial neural nets are able to account for violations of conditional independence and yield very similar results. Weights-of-evidence cannot reasonably include interaction terms; subsequent modifications of the weights, as often suggested, cannot emulate the effect of interaction terms.
ISPRS Int. J. Geo-Inf.2014, 3(4), 1372-1386; doi:10.3390/ijgi3041372 - published 10 December 2014 Show/Hide Abstract
Abstract: The NASA Giovanni data analysis system has been recognized as a useful tool to access and analyze many different types of remote sensing data. The variety of environmental data types has allowed the use of Giovanni for different application areas, such as agriculture, hydrology, and air quality research. The use of Giovanni for researching connections between public health issues and Earth’s environment and climate, potentially exacerbated by anthropogenic influence, has been increasingly demonstrated. In this communication, the pertinence of several different data parameters to public health will be described. This communication also provides a case study of the use of remote sensing data from Giovanni in assessing the associations between seasonal influenza and meteorological parameters. In this study, logistic regression was employed with precipitation, temperature and specific humidity as predictors. Specific humidity was found to be associated (p < 0.05) with influenza activity in both temperate and tropical climate. In the two temperate locations studied, specific humidity was negatively correlated with influenza; conversely, in the three tropical locations, specific humidity was positively correlated with influenza. Influenza prediction using the regression models showed good agreement with the observed data (correlation coefficient of 0.5–0.83).
ISPRS Int. J. Geo-Inf.2014, 3(4), 1352-1371; doi:10.3390/ijgi3041352 - published 10 December 2014 Show/Hide Abstract
Abstract: Controlling dengue virus transmission mainly involves integrated vector management. Risk maps at appropriate scales can provide valuable information for assessing entomological risk levels. Here, results from a spatio-temporal model of dwellings potentially harboring Aedes aegypti larvae from 2009 to 2011 in Tartane (Martinique, French Antilles) using high spatial resolution remote-sensing environmental data and field entomological and meteorological information are presented. This tele-epidemiology methodology allows monitoring the dynamics of diseases closely related to weather/climate and environment variability. A Geoeye-1 image was processed to extract landscape elements that could surrogate societal or biological information related to the life cycle of Aedes vectors. These elements were subsequently included into statistical models with random effect. Various environmental and meteorological conditions have indeed been identified as risk/protective factors for the presence of Aedes aegypti immature stages in dwellings at a given date. These conditions were used to produce dynamic high spatio-temporal resolution maps from the presence of most containers harboring larvae. The produced risk maps are examples of modeled entomological maps at the housing level with daily temporal resolution. This finding is an important contribution to the development of targeted operational control systems for dengue and other vector-borne diseases, such as chikungunya, which is also present in Martinique.