In surveillance and monitoring systems, the use of mobile vehicles or unmanned aerial vehicles (UAVs), like the drone type, provides advantages in terms of access to the environment with enhanced range, maneuverability, and safety due to the ability to move omnidirectionally to explore, identify, and perform some security tasks. These activities must be performed autonomously by capturing data from the environment; usually, the data present errors and uncertainties that impact the recognition and resolution in the detection and identification of objects. The resolution in the acquisition of data can be improved by integrating data sensor fusion systems to measure the same physical phenomenon from two or more sensors by retrieving information simultaneously. This paper uses the constant turn and rate velocity (CTRV) kinematic model of a drone but includes the angular velocity not considered in previous works as a complementary alternative in Lidar and Radar data sensor fusion retrieved using UAVs and applying the extended Kalman filter (EKF) for the detection of moving targets. The performance of the EKF is evaluated by using a dataset that jointly includes position data captured from a LiDAR and a Radar sensor for an object in movement following a trajectory with sudden changes. Additive white Gaussian noise is then introduced into the data to degrade the data. Then, the root mean square error (RMSE) versus the increase in noise power is evaluated, and the results show an improvement of 0.4 for object detection over other conventional kinematic models that do not consider significant trajectory changes.

This paper evaluates the performance of spectrum sensing techniques like energy detector and matched filter in Cognitive Radio systems. To solve the problem of spectral detection two hypotheses H0 and H1, are considered. A statistic test over the received data is compared with a predetermined threshold, to check the availability of the channel and the accuracy of the detector. Hence, a detection probability (PD) and false alarm probability (PFA) are defined for a specific signal to noise ratio (SNR). A simulation is performed in Matlab and is validated on GNURadio using a board USRP N210 to detect frequency modulated (FM) signal at 100 MHz. The receiver operating characteristic (ROC) curves are used to illustrate the performance of binary classifier system when its discrimination threshold is modified. The matched filter gets a better performance than the energy detector under some characteristics and agrees with the requirements established to changes in PFA and noise power, besides has a higher measurement range (signal detection allows -25 dB SNR).

This paper presents a novel methodology for deploying wireless sensor nodes in the Industrial Internet of Things (IIoT) to address the safety and efficiency challenges in underground coal mining. The methodology is intended to support long-term planning on mitigating the risks in occupational health and safety policies. To ensure realistic and accurate deployment, we propose a software tool that generates mine models based on geolocation data or blueprints in image format, allowing precise adaptation to the specific conditions of each mine. Furthermore, the process is based on sensing and communication range values obtained through simulations and on-site experiments. The deployment strategy is articulated in two complementary steps: a deterministic deployment, where nodes are strategically placed according to the structure of the tunnels, followed by a random stage to include additional nodes that ensure optimal coverage and connectivity inside the mine by comparing different methodologies for deploying sensor networks using coverage density as a performance metric. We analyze coverage and connectivity based on the three probability density functions (PDFs) for the random deployment of nodes: uniform, normal, and exponential, evaluating both the degree of coverage (k-coverage) and the degree of connectivity (k-connectivity). The results show that our proposed methodology stands out for its lower density of sensors per square meter, which translates into a reduction of between 20.81% and 23.46% for uniform and exponential PDFs, respectively, concerning the number of sensors compared to the analyzed methodologies. In this way, it is possible to determine which distribution is suitable to cover the elongated area with the smallest number of nodes, considering the coverage and connectivity requirements, to reduce the deployment cost. The uniform PDF minimizes the number of sensors needed by 44.70% in small mines and 46.27% in medium ones compared to the exponential PDF. These findings provide valuable information to optimize node deployment regarding cost and efficiency; a uniform function is a good option depending on prices. The exponential distribution reached the highest values of k-coverage and k-connectivity for small and medium-sized mines; in addition, it has greater robustness and tolerance to faults like signal network intermittence. This methodology not only improves the collection of critical information for the mining operation but also plays a vital role in reducing the risks to the health and safety of workers by providing a more robust and adaptive monitoring system. The approach can be used to plan IIoT systems based on Wireless Sensor Networks (WSN) for underground mining exploitation, offering a more reliable and adaptable strategy for monitoring and managing complex work environments.

La presencia del tizón tardío o gota en el cultivo de papa afecta directamente el crecimiento de la planta y el desarrollo del tubérculo, por ello, es importante la detección temprana de la enfermedad. Actualmente, la aplicación de redes neuronales convolucionales es una oportunidad orientada a la identificación de patrones en la agricultura de precisión, incluyendo el estudio del tizón tardío, en el cultivo de papa. Este estudio describe un modelo de aprendizaje profundo capaz de reconocer el tizón tardío en el cultivo de papa, por medio de la clasificación de imágenes de las hojas. Se utilizó, en la aplicación de este modelo, el conjunto de datos aumentado de PlantVillage, para entrenamiento. El modelo propuesto ha sido evaluado a partir de métricas de rendimiento, como precisión, sensibilidad, puntaje F1 y exactitud. Para verificar la efectividad del modelo en la identificación y la clasificación del tizón tardío y comparado en rendimiento con arquitecturas. como AlexNet, ZFNet, VGG16 y VGG19. Los resultados experimentales obtenidos con el conjunto de datos seleccionado mostraron que el modelo propuesto alcanza una exactitud del 90 % y un puntaje F1, del 91 %. Por lo anterior, se concluye que el modelo propuesto es una herramienta útil para los agricultores en la identificación del tizón tardío y escalable a plataformas móviles, por la cantidad de parámetros que lo comprenden.

The use of LiDAR technologies offer an effective alternative for enhancing monitoring and surveillance tasks in secured areas, or hard-to-access areas. Conventional solutions allow detection and identification capabilities, wich are valuable for object and pattern recognition in environments with special conditions. This paper proposes a method for structure a dataset by collecting data from LiDAR sensors and reconstructing 3D point cloud scenes into images. Furthermore, training with these data using semantic segmentation algorithms allows for the detecting of people in monitoring and surveillance contexts. The process involves data acquisition techniques, preprocessing, and the application of tools and technologies for semantic segmentation in the post-processing stage. Data is acquired from RPLidar sensor with a spatial resolution of less than 0.4 degrees per step, ensuring greater accuracy in detecting small objects. Scanning distances between 0.15 and 12 meters allow for the capture of 941 points per rotation angle. The dataset is complemented with labels and images containing binary masks of regions of interest for people detection using semantic segmentation. The main contribution of this paper is the methodological description of constructing a dataset comprising 1,254 images in various scenarios, including indoors and outdoors, with different levels of occlusion and illumination. This dataset facilitates the development of machine learning algorithms and deep learning algorithms for pattern recognition, such as Mask R-CNN (89.9% mean class accuracy) and U-NET (90.5% mean class accuracy). The results demonstrated an approximate 2% improvement in mean class accuracy compared to other state-of-the-art algorithms in semantic segmentation for point clouds.