Digital Precision Agriculture (DPA) is a comprehensive approach to agronomic management that utilizes advanced technologies, such as sensor data analysis and automation, to optimize crop productivity, enhance farm income, and minimize environmental impacts. DPA encompasses various agricultural domains, including pest control, pest management, fertilization, irrigation management, sowing, transplanting, crop health monitoring, yield forecasting, harvesting, and post-harvest stages. Among the enabling technologies for DPA, Unmanned Aerial Vehicles (UAVs) have gained significant attention and market growth. The advancements in control systems, robotics, electronics, and artificial intelligence have led to the development of sophisticated agricultural drones. UAVs offer advantages such as versatility, quick and accurate remote sensing capabilities, and high-quality imaging at affordable prices. Furthermore, the miniaturization of sensors and advancements in nanotechnology enable UAVs to perform multiple operations simultaneously without compromising flight autonomy. However, various variables, including aircraft mass, payload capacity, size, battery characteristics, flight autonomy, cost, and environmental conditions, impact the performance and applicability of UAV systems in agriculture. The economic considerations involve the purchase of drones, equipment, and the expertise of trained pilots for flight management and data processing. Payload capacity, flight range, and financial factors influence the choice and implementation of UAVs in agriculture. The research and patent trends show the growing interest in UAVs for agricultural applications. This paper provides a general review of UAV types, construction architectures, and their diverse applications in agriculture until 2022.

Advanced intelligent surveillance systems are able to automatically analyze video of surveillance data without human intervention. These systems allow high accuracy of human activity recognition and then a high-level activity evaluation. To provide such features, an intelligent surveillance system requires a background subtraction scheme for human segmentation that captures a sequence of images containing moving humans from the reference background image. This paper proposes an alternative approach for human segmentation in videos through the use of a deep convolutional neural network. Two specific datasets were created to train our network, using the shapes of 35 different moving actors arranged on background images related to the area where the camera is located, allowing the network to take advantage of the entire site chosen for video surveillance. To assess the proposed approach, we compare our results with an Adobe Photoshop tool called Select Subject, the conditional generative adversarial network Pix2Pix, and the fully-convolutional model for real-time instance segmentation Yolact. The results show that the main benefit of our method is the possibility to automatically recognize and segment people in videos without constraints on camera and people movements in the scene (Video, code and datasets are available at http://graphics.unibas.it/www/HumanSegmentation/index.md.html).

Today, virtual reality (VR) systems are widely available through low-cost devices such as Oculus Rift and HTC Vive. Although VR technology has so far been centered on entertainment, there is a growing interest from developers, technology companies, and consumers to evaluate it in a wider variety of contexts. This paper explores the effectiveness of visualizing and interacting with three-dimensional graphs in VR in comparison with the traditional approach. In particular, we present an empirical evaluation study for exploring and interacting with three-dimensional graphs using Oculus Rift and Leap Motion. We designed several interfaces exploiting the natural user interface in a VR environment and compared them with traditional mouse–keyboard and joypad configurations. Our evaluation suggests that, although these upcoming VR technologies are more challenging than more traditional ones, they facilitate user involvement during graph interaction and visualization tasks, given the enjoyable experience elicited when combining gesture-based interfaces and VR.

Virtual reality has achieved significant popularity in recent years, and allowing users to move freely within an immersive virtual world has become an important factor critical to realize. The user’s interactions are generally designed to increase the perceived realism, but the locomotion techniques and how these affect the user’s task performance still represent an open issue, much discussed in the literature. In this article, we evaluate the efficiency and effectiveness of, and user preferences relating to, freehand locomotion techniques designed for an immersive virtual environment performed through hand gestures tracked by a sensor placed in the egocentric position and experienced through a head-mounted display. Three freehand locomotion techniques have been implemented and compared with each other, and with a baseline technique based on a controller, through qualitative and quantitative measures. An extensive user study conducted with 60 subjects shows that the proposed methods have a performance comparable to the use of the controller, further revealing the users’ preference for decoupling the locomotion in sub-tasks, even if this means renouncing precision and adapting the interaction to the possibilities of the tracker sensor.