Abstract: Spatial effects are intrinsic to studies on travel behavior, including modal choice problems. Among spatial dependence and heterogeneity, the resulting spatial variation of estimated coefficients is one of the most significant gains in local spatial models. Thus, this article aims to analyze a modal choice problem, considering spatial dependence and variability simultaneously. Additionally, a test is proposed to evaluate and validate the spatial variability, obtaining disaggregated results. A database adapted from the household origin-destination survey was used, which was carried out in 2007/2008 in the city of São Carlos - São Paulo, Brazil. The proposed spatial variability test uses the estimated parameters of the GWLR model (main database and 200 spatially randomized databases), compared to the confidence intervals of the coefficients in the non-spatial logit model for the spatial variability hypothesis. The results of the proposed test are similar to the reference test in the case study. The disaggregated results can be used to verify if there are certain subgroups that are more likely to be spatially stationary and if these groups exhibit a spatial pattern. Moreover, it can be observed that the local spatial model provides a better fit and estimates when compared to the non-spatial model.

Abstract: The present paper addresses the relevance of incorporating terrain data for analyzing satellite images in mapping land use and land cover in the Cerrado biome. Assuming that terrain influences the dynamics of landscape changes, the present investigation evaluates three machine learning algorithms: Random Forest (RF), Decision Tree (DT), and Support Vector Machine (SVM) in a watershed with significant topographic heterogeneity. The present study evaluated variations in image classification using Sentinel-2 satellite data. It also included a composite analysis blending Sentinel-2 data and information derived from the Shuttle Radar Topography Mission (SRTM). The results indicate that SVM exhibited the best performance, both with and without terrain data. Although DT demonstrated satisfactory results, the performance was inferior to SVM. However, DT´s significantly shorter processing time presents an advantage in scenarios involving large territorial extents or computational constraints. Conversely, RF had a processing time similar to DT but recorded the lowest statistical indices among the three algorithms. Additionally, including the data cube containing elevation data and its derivatives yielded improved land use and land cover classification results for all evaluated algorithms compared to images without terrain data. This demonstrates the robustness of the process and the significant improvement in the quality of the final product.

Abstract: The term City Information Modeling - CIM is a recent concept that encompasses, among other aspects, the integration between Building Information Modeling and Geographic Information Systems. CIM has been predominantly applied in cities; however, since the environment of a university campus possesses characteristics similar to a municipality, CIM can be a promising tool for managing these institutions. This paper proposes a data organization model for CIM, based on CityGML extensions. A methodology was proposed for the creation of extensions that can be applied to activities requiring spatial and building data. As an example, the activity of developing a technical project for fire and disaster prevention was used, in the environment of the Federal University of Paraná. To create the method, the available data and its format were studied, as well as its geometric and semantic correspondence with CityGML. The result was a conceptual model for the application domain extension (ADE) of CityGML for the development of firefighting projects, referred to in this work as ADE_FirePrev.

Abstract: Updating maps in Brazil is hindered by considerable obstacles, primarily to the high costs associated with it and the difficulty of accessing the regions. Moreover, the accelerated rate of environmental transformation, particularly in rural settings, represents an additional challenge. This study proposes to use high-resolution satellite images from Planet constellation, in conjunction with artificial intelligence, specifically UNet, to automatically identify rural roads in the metropolitan region of Curitiba, Paraná, Brazil. The objective is to identify the optimal parameters for automating the detection of rural roads. The UNet, with its distinctive U-shaped architecture, is highly effective in segmenting and detecting targets while simultaneously preserving the feature maps in each convolution. In this study, the network was trained on satellite images containing rural roads, resulting in segmented maps with an encouraging 91.95% accuracy in road detection. Nevertheless, further improvement is possible, as evidenced by the method’s precision of 75.83% and F1-Score of 69.07%. These outcomes indicate the possibility of enhancement through the expansion of the training dataset, thereby better addressing the network’s recognition constraints. One potential avenue for optimizing detection using the methodology would be the incorporation of supplementary training samples, which could potentially mitigate the network’s recognition limitations.

Abstract: This research aims to evaluate the preference for cartographic symbol designs that depict urban buildings for topographic maps at a 1:10,000 scale, to define symbols that may represent part of the western urban landscape of the state of São Paulo. Three proposals were designed to represent buildings: Pink visually coherent with the national cartographic standards; Orange visually coherent with the urban landscape of the western portion of São Paulo state; Gray visually coherent with the digital navigational maps used frequently. Map evaluation considered the preference for the symbol designs between architects and non-architects. The Orange design was the most preferred in all aspects. Despite that, Orange and Pink designs did not present statistically significant differences, but Gray was the least preferred. Furthermore, from the statistical analysis, map users with a background in architecture and those with other backgrounds could be considered as part of the same population.

Abstract: The determination of the position and velocity of the GLONASS user’s begins with the GLONASS satellite’s orbit computation. In this paper, using the positions and velocities of GLONASS satellites given in the broadcast ephemerides file every 30 minutes as initial conditions, we study the effect of the order and the step size in the integration of the differential equation of satellite motion by the Runge-Kutta method to get the positions and velocities of GLONASS satellites at any time. This method consists of several orders, including order four recommended in the GLONASS Interface Control Document; order five; and order four and five. These three orders are tested in this study using two distinct step sizes (1 sec and 0.5 sec). In terms of the differences obtained between the forward (+15 min) and the backward (-15 min) integration processes and the runtime, order four is the most suitable for the determination of GLONASS orbits compared to the other orders employed in this study. The data used is the positions, velocities, and luni-solar acceleration of all GLONNAS satellites on April 20, 2021, given in the broadcast ephemerides file.

Abstract: Map evaluation is a fundamental and challenging aspect of cartography. This is especially true for digital maps that can be accessed via mobile devices (hereafter referred to as «mobile digital maps»). Researchers often use a combination of methods adapted from computer science to evaluate these maps in order to obtain comprehensive and reliable information. However, the evaluation of digital maps is a complex task that requires a multidimensional approach to understand the different perspectives and needs of users in different usage contexts. By analyzing scientific articles, this study aims to explore and identify the most commonly used methods for evaluating mobile digital maps. The goal is to provide an analytical review of the methods, test configurations, and application locations of mobile digital map evaluations. For this research, 195 scientific articles published in international journals between 2010 and 2023 were evaluated. Of these, 53 articles were selected, all of which dealt with the evaluation of digital maps accessed by mobile devices. The study answers the question of which methods are recommended depending on the context. The types of data collected cover both qualitative and quantitative aspects, while the evaluations tend to be moderate in nature.

Abstract: Multipath significantly degrades GNSS positioning quality, especially in smartphones due to their inferior GNSS antennas. To mitigate this, the Laboratory of Space Geodesy and Hydrography (LAGEH) developed the Multipath Effect Attenuator (AEM), which was used for the first time in smartphone applications. A new prototype, AEM Smart, was also created for use in smartphones. The smartphone used in this research was the Xiaomi Mi 8, with a dual frequency GNSS sensor and allows access to raw GNSS data. The multipath effect was evaluated based on the MP1 and MP5 indices, referring to the L1 and L5 carrier phases, respectively. The positioning quality was assessed using Precise Point Positioning (PPP) GNSS data processing. This contribution shows that L5 is significantly superior to L1 measurements, with more than 73% improvement in multipath effect suppression. Furthermore, this study indicates that attenuators can reduce the multipath effect in L1 by more than 24%. Also, a strong correlation of about 94% was found between the multipath effect on L1 and the 2D and 3D positional accuracies. Finally, through the PPP technique in combination with the Xiaomi Mi 8 and AEM 3, it was possible to obtain a 2D and 3D accuracy of approximately 0.24 and 0.60 m, respectively.

Abstract: The socio-environmental impacts caused by the collapse of a mining dam can be irreversible. In Brazil, the dam collapse at the Córrego do Feijão Mine was considered one of the worst disasters in the country. Remote sensing-based approaches have been used to detect and monitor areas affected by tailings from dam rupture. Therefore, it was proposed to identify the area affected by mining tailing mud, in Brumadinho, Minas Gerais State, through the analysis of three different spectral indices: Normalized Difference Vegetation Index (NDVI), Ferrous Minerals Ratio (FMR) and Clay Minerals Ratio (CMR). These indices were computed from the Operational Land Imager (OLI) images of Landsat-8. Different thresholds were tested to define the best range for delineating the affected area. For validation, the limits of the affected area, obtained from a higher resolution sensor, GeoEye-1, were used as reference. The methodology demonstrated great potential for detecting areas affected by dam failure. The indices NDVI and FMR delimited the area of interest with high performance, with precision varying between 95% and 92%; recall between 88% and 87%; F-score between 91% and 89%; and global accuracy between 84% and 80%, showing to be suitable mapping such disasters.

Abstract: Groundwater resources, a crucial water supply in the Sahel regions of northern Africa, have to be evaluated to ensure its sustainability. In the last decade, Lake Chad Basin (LCB), experienced extreme variabilities in the surface water, which impacted groundwater levels as studied by different authors. Consequently, evaluating the modes of variability of each compartment of the terrestrial water storage (TWS) plays an essential role in LCB water resources management. Therefore, the main objective of this study is to invert groundwater storage (GWS) over LCB using surface water storage (SWS) derived from satellite altimetry and imagery, soil moisture storage (SMS) from the Global Land Data Assimilation System hydrological model (GLDAS), and TWS inferred from Gravity Recovery and Climate Experiment (GRACE) measurements. Here we apply a proper signal separation approach to extract GWS from TWS, which is taken as observations in a least-squares sense to reconstruct GWS. The analysis of the reconstructed GWS series using correlation coefficient, root-mean-square error, and the Nash-Sutcliffe efficiency, shows an improvement of 114%, 26%, and 147%, respectively, compared to the GRACE-based results. These findings could be crucial in the surface and sub-surface water management over LCB.

Abstract: Urban infrastructure element detection is important for the domain of public management in large urban centres. The diversity of objects in the urban environment makes object detection and classification a challenging task, requiring fast and accurate methods. Advances in deep learning methods have driven improvement in detection techniques (processing, speed, accuracy) that do not rely on manually crafted models, but, instead, use learning approaches with corresponding large training sets to detect and classify objects in images. We applied an object detection model to identify and classify four urban infrastructure elements in the Mappilary dataset. We use YOLOv5, one of the top-performing object detection models, a recent release of the YOLO family, pre-trained on the COCO dataset but fine-tuned on Mappilary dataset. Experimental results from the dataset show that YOLOv5 can make qualitative predictions, for example, the power grid pole class presented the mean Average Precision (mAP) of 78% and the crosswalk class showed mAP around 79%. A lower degree of certainty was verified in the detection of public lighting (mAP=64%) and accessibility (mAP=61%) classes due to the low resolution of certain objects. However, the proposed method showed the capability of automatically detection and location of urban infrastructure elements in real-time, which could contribute to improve decision-making.

Abstract: Mapping the submerged bottom is a hampered task when traditional vessels are inserted in shallow places that pose a danger to navigation. In this sense, current research has sought optical remote sensing to obtain bathymetry estimates in larger locations in less time. However, most studies employ a large sample of bathymetric points to estimate depth with orbital images. The need for a large amount of random bathymetric points can make these procedures less viable and unattractive. Thus, the present work proposes a methodology to estimate bathymetric depths from orbital images using sections of bathymetric points previously spaced in the study area, avoiding the need for a high amount of points collected through traditional bathymetric surveys. This work also compares this methodology using the NDWI index and the ANNs. Furthermore, the study showed that the points contained in the sections are quite efficient for extracting bathymetry with orbital images, especially through the implementation of neural networks, achieving a volume estimation accuracy within 4% of the actual volume of the reservoir in question.

Abstract: Tomatoes are widely cultivated, both by family farmers and corporate producers. During the tomato growth cycle, several diseases can affect the plant. The identification of these diseases through short-range images is significant, and computer vision techniques are commonly used to identify diseases in plant leaves. In this paper, a hybrid model that combines a convolutional neural network (CNN) and a Random Forest (RF) decision tree is used for foliar spot detection in tomato leaves. High-level features learned and extracted from CNN are used as input for the RF classifier. To evaluate the proposed model’s performance for plant disease identification, a case study of 2480 low-cost digital RGB images collected in actual field conditions, under different intensities of light exposure, were used, including healthy tomato leaves and leaves with visible symptoms of powdery mildew fungus, which attacks the tomato leaf. The results were compared with six conventional machine learning classifiers: Logistic Regression (LR), Linear Discriminant Analysis (LDA), K- Nearest Neighbors (KNN), Naive Bayes (NB), Support Vector Machine (SVM) and Random Forest (RF). The results show that the proposed model outperformed conventional classifiers, reaching an accuracy of 98%. The results highlight the importance of fusing models to improve the detection plant´s diseases.

Abstract: Pan-sharpening (PS) consists of combining a high spatial resolution (HR) panchromatic image (PAN) and a low spatial resolution (LR) multispectral image (MS) to generate an MS-HR image. However, some PS methods have spectral and spatial distortions that influence subsequent analyses. Thus, this study aimed to develop a PS method based on convolutional autoencoder (CAE) for Landsat 8 images and evaluate its performance in calculating spectral indices. In the PS process, we trained a CAE network and used a multiscale-guided filter. The performance of the proposed method was analyzed using the Kolmogorov-Smirnov (K-S) statistic of the empirical cumulative distribution function (eCDF) between the values of the Normalized Difference Vegetation Index (NDVI), Normalized Difference Water Index (NDWI), and Normalized Difference Moisture Index (NDMI) of the PS and MS-LR images. The results show that the proposed method is effective for calculating the indices. Therefore, we conclude that it has great potential for preserving the spatial information of the PAN image and the spectral information of the MS-LR image during the PS process for calculating spectral indices.

Abstract: Augmented Reality (AR) seamlessly integrates the real environment with virtual objects, enriching users’ information perception. The application of AR in personal and vehicular navigation tasks facilitates the evolution of navigation systems from two-dimensional (2D) to three-dimensional (3D) representations, incorporating an egocentric viewpoint. Spatial knowledge, specific navigation objectives, navigational skills, and the navigational symbols plays pivotal roles in map utilization, encompassing self-localization, proximity, navigation, and event awareness. The representation of symbols in navigation tasks relies on static and dynamic or animated visual cartographic variables. In this context, our research focuses on evaluating symbols used for depicting routes and landmarks in AR systems during personal navigation tasks. The experiments investigated the representation of routes, exploring variations in route line thickness, the presence or absence of border-color, fill color, and the speed of arrows moving along the route line. The findings of the study reveal that thinner route lines significantly contribute to enhance perception of surroundings, blue fill color exhibit superior performance compared to traditional navigation system colors, and volunteers noted that the arrow animation speed noticeably affected their perceptions of elements within the scenario. These insights contribute to understanding of how visual variables influence user preferences and experiences in AR-based navigation tasks.