Search

Land cover classification, generated from satellite imagery through semantic segmentation, has become fundamental for monitoring land use and land cover change (LULCC). The tropical Andes territory provides opportunities due to its significance in the provision of ecosystem services. However, the lack of reliable data for this region, coupled with challenges arising from its mountainous topography and diverse ecosystems, hinders the description of its coverage. Therefore, this research proposes the Tropical Andes Land Cover Dataset (TALANDCOVER). It is constructed from three sample strategies: aleatory, minimum 50%, and 70% of representation per class, which address imbalanced geographic data. Additionally, the U-Net deep learning model is applied for enhanced and tailored classification of land covers. Using high-resolution data from the NICFI program, our analysis focuses on the Department of Antioquia in Colombia. The TALANDCOVER dataset, presented in TIF format, comprises multiband R-G-B-NIR images paired with six labels (dense forest, grasslands, heterogeneous agricultural areas, bodies of water, built-up areas, and bare-degraded lands) with an estimated 0.76 F1 score compared to ground truth data by expert knowledge and surpassing the precision of existing global cover maps for the study area. To the best of our knowledge, this work is a pioneer in its release of open-source data for segmenting coverages with pixel-wise labeled NICFI imagery at a 4.77 m resolution. The experiments carried out with the application of the sample strategies and models show F1 score values of 0.70, 0.72, and 0.74 for aleatory, balanced 50%, and balanced 70%, respectively, over the expert segmented sample (ground truth), which suggests that the personalized application of our deep learning model, together with the TALANDCOVER dataset offers different possibilities that facilitate the training of deep architectures for the classification of large-scale covers in complex areas, such as the tropical Andes. This advance has significant potential for decision making, emphasizing sustainable land use and the conservation of natural resources.

Vegetation encroachment in power line corridors remains a major challenge for modern energy-dependent societies, as it can cause power outages and lead to significant financial losses. Unmanned Aerial Vehicles (UAVs) have emerged as a promising solution for monitoring infrastructure, owing to their ability to acquire high-resolution overhead images of these areas quickly and affordably. However, accurate segmentation of the vegetation encroachment in this imagery is a challenging task, due to the complexity of the scene and the high pixel imbalance between the power lines, the vegetation and the background classes. In this paper, we propose a deep learning-based approach to tackle this problem caused by the original and different geometry of the objects. Specifically, we use DeepLabV3, U-Net and a modified version of the U-Net architecture with VGG-16 weights to train two separate models. One of them segments the dominant classes, the vegetation from the background, achieving an IoU of 0.77. The other one segments power line corridors from the background, obtaining an IoU of 0.64. Finally, ensembling both models into one creates an “encroachment” zone, where power lines and vegetation are intersected. We train our models using the Vegetation Encroachment in Power Line Corridors dataset (VEPL), which includes RGB orthomosaics and multi-label masks for segmentation. Experimental results demonstrate that our approach outperforms individual networks and original prominent architectures when applied to this specific problem. This approach has the potential to significantly improve the efficiency and accuracy of vegetation encroachment monitoring using UAV, thus helping to ensure the reliability and sustainability of power supply.

The COVID-19 pandemic has had a significant impact on various aspects of society, including economic, health, political, and work-related domains. The pandemic has also caused an emotional effect on individuals, reflected in their opinions and comments on social media platforms, such as Twitter. This study explores the evolution of sentiment in Spanish pandemic tweets through a data analysis based on a fine-tuned BERT architecture. A total of six million tweets were collected using web scraping techniques, and pre-processing was applied to filter and clean the data. The fine-tuned BERT architecture was utilized to perform sentiment analysis, which allowed for a deep-learning approach to sentiment classification. The analysis results were graphically represented based on search criteria, such as “COVID-19” and “coronavirus”. This study reveals sentiment trends, significant concerns, relationship with announced news, public reactions, and information dissemination, among other aspects. These findings provide insight into the emotional impact of the COVID-19 pandemic on individuals and the corresponding impact on social media platforms.

Drone imagery is becoming the main source of overhead information to support decisions in many different fields, especially with deep learning integration. Datasets to train object detection and semantic segmentation models to solve geospatial data analysis are called GeoAI datasets. They are composed of images and corresponding labels represented by full-size masks typically obtained by manual digitizing. GIS software is made of a set of tools that can be used to automate tasks using geo-referenced raster and vector layers. This work describes a workflow using GIS tools to produce GeoAI datasets. In particular, it mentions the steps to obtain ground truth data from OSM and use methods for geometric and spectral augmentation and the data fusion of drone imagery. A method semi-automatically produces masks for point and line objects, calculating an optimum buffer distance. Tessellation into chips, pairing and imbalance checking is performed over the image–mask pairs. Dataset splitting into train–validation–test data is done randomly. All of the code for the different methods are provided in the paper, as well as point and road datasets produced as examples of point and line geometries, and the original drone orthomosaic images produced during the research. Semantic segmentation results performed over the point and line datasets using a classical U-Net show that the semi-automatically produced masks, called primitive masks, obtained a higher mIoU compared to other equal-size masks, and almost the same mIoU metric compared to full-size manual masks.

Detection and Semantic Segmentation of vehicles in drone aerial orthomosaics has applications in a variety of fields such as security, traffic and parking management, urban planning, logistics, and transportation, among many others. This paper presents the HAGDAVS dataset fusing RGB spectral channel and Digital Surface Model DSM for the detection and segmentation of vehicles from aerial drone images, including three vehicle classes: cars, motorcycles, and ghosts (motorcycle or car). We supply DSM as an additional variable to be included in deep learning and computer vision models to increase its accuracy. RGB orthomosaic, RG-DSM fusion, and multi-label mask are provided in Tag Image File Format. Geo-located vehicle bounding boxes are provided in GeoJSON vector format. We also describes the acquisition of drone data, the derived products, and the workflow to produce the dataset. Researchers would benefit from using the proposed dataset to improve results in the case of vehicle occlusion, geo-location, and the need for cleaning ghost vehicles. As far as we know, this is the first openly available dataset for vehicle detection and segmentation, comprising RG-DSM drone data fusion and different color masks for motorcycles, cars, and ghosts.

In recent years, the use of deep learning-based models for developing advanced healthcare systems has been growing due to the results they can achieve. However, the majority of the proposed deep learning-models largely use convolutional and pooling operations, causing a loss in valuable data and focusing on local information. In this paper, we propose a deep learning-based approach that uses global and local features which are of importance in the medical image segmentation process. In order to train the architecture, we used extracted three-dimensional (3D) blocks from the full magnetic resonance image resolution, which were sent through a set of successive convolutional neural network (CNN) layers free of pooling operations to extract local information. Later, we sent the resulting feature maps to successive layers of self-attention modules to obtain the global context, whose output was later dispatched to the decoder pipeline composed mostly of upsampling layers. The model was trained using the Mindboggle-101 dataset. The experimental results showed that the self-attention modules allow segmentation with a higher Mean Dice Score of 0.90 ± 0.036 compared with other UNet-based approaches. The average segmentation time was approximately 0.038 s per brain structure. The proposed model allows tackling the brain structure segmentation task properly. Exploiting the global context that the self-attention modules incorporate allows for more precise and faster segmentation. We segmented 37 brain structures and, to the best of our knowledge, it is the largest number of structures under a 3D approach using attention mechanisms.

According to worldwide statistics, traffic accidents are the cause of a high percentage of violent deaths. The time taken to send the medical response to the accident site is largely affected by the human factor and correlates with survival probability. Due to this and the wide use of video surveillance and intelligent traffic systems, an automated traffic accident detection approach becomes desirable for computer vision researchers. Nowadays, Deep Learning (DL)-based approaches have shown high performance in computer vision tasks that involve a complex features relationship. Therefore, this work develops an automated DL-based method capable of detecting traffic accidents on video. The proposed method assumes that traffic accident events are described by visual features occurring through a temporal way. Therefore, a visual features extraction phase, followed by a temporary pattern identification, compose the model architecture. The visual and temporal features are learned in the training phase through convolution and recurrent layers using built-from-scratch and public datasets. An accuracy of 98% is achieved in the detection of accidents in public traffic accident datasets, showing a high capacity in detection independent of the road structure.

High-resolution 3D scanning devices produce high-density point clouds, which require a large capacity of storage and time-consuming processing algorithms. In order to reduce both needs, it is common to apply surface simplification algorithms as a preprocessing stage. The goal of point cloud simplification algorithms is to reduce the volume of data while preserving the most relevant features of the original point cloud. In this paper, we present a new point cloud feature-preserving simplification algorithm. We use a global approach to detect saliencies on a given point cloud. Our method estimates a feature vector for each point in the cloud. The components of the feature vector are the normal vector coordinates, the point coordinates, and the surface curvature at each point. Feature vectors are used as basis signals to carry out a dictionary learning process, producing a trained dictionary. We perform the corresponding sparse coding process to produce a sparse matrix. To detect the saliencies, the proposed method uses two measures, the first of which takes into account the quantity of nonzero elements in each column vector of the sparse matrix and the second the reconstruction error of each signal. These measures are then combined to produce the final saliency value for each point in the cloud. Next, we proceed with the simplification of the point cloud, guided by the detected saliency and using the saliency values of each point as a dynamic clusterization radius. We validate the proposed method by comparing it with a set of state-of-the-art methods, demonstrating the effectiveness of the simplification method.

Denoising the point cloud is fundamental for reconstructing high quality surfaces with details in order to eliminate noise and outliers in the 3D scanning process. The challenges for a denoising algorithm are noise reduction and sharp features preservation. In this paper, we present a new model to reconstruct and smooth point clouds that combine L1-median filtering with sparse L1 regularization for both denoising the normal vectors and updating the position of the points to preserve sharp features in the point cloud. The L1-median filter is robust to outliers and noise compared to the mean. The L1 norm is a way to measure the sparsity of a solution, and applying an L1 optimization to the point cloud can measure the sparsity of sharp features, producing clean point set surfaces with sharp features. We optimize the L1 minimization problem by using the proximal gradient descent algorithm. Experimental results show that our approach is comparable to the state-of-the-art methods, as it filters out 3D models with a high level of noise, but keeps their geometric features.

With the advancements in neuroimaging techniques, understanding the relationship between brain morphology and behavioral tendencies such as criminal behavior has garnered interest. This research addresses the investigation of disparities in neuroanatomical structures between adolescent offenders and non-offenders and considers the implications of such distinctions regarding offender behavior within adolescent populations. Employing data-driven methodologies, MRI scans of adolescents from Barranquilla, Colombia, were analyzed to explore morphological variations. Utilizing a 1.5 Tesla Siemens resonator (Siemens Healthineers, Erlangen, Germany), T1-weighted MPRAGE anatomical images were acquired and analyzed using a systematic five-step methodology including data acquisition, MRI pre-processing, feature selection, model selection, and model validation and evaluation. Participants, both offenders and non-offenders, were aged 14–18 and selected based on education, criminal history, and physical conditions. The research identified significant disparities in the volumes of 42 brain structures between adolescent offenders (AOs) and non-offenders (NOs), highlighting particular brain regions potentially associated with offending behavior. Additionally, a considerable proportion of AOs emanated from lower socioeconomic backgrounds and showcased marked substance use. The findings suggest that neuroanatomical disparities potentially correlate with criminal behavior among adolescents at a neurobiological level. Noticeable socio-environmental factors, such as lower socioeconomic status and substance abuse, were substantially prevalent among AOs. Particularly, neurobiological deviations in structures like ctx-lh-rostralmiddlefrontal and ctx-lh-caudalanteriorcingulate perhaps represent a link between neurological factors and external stimuli.