Super-linear speedup for real-time condition monitoring using image processing and drones

Real-time inspections for the large-scale solar system may take a long time to get the hazard situations for any failures that may take place in the solar panels normal operations, where prior hazards detection is important. Reducing the execution time and improving the system’s performance are the ultimate goals of multiprocessing or multicore systems. Real-time video processing and analysis from two camcorders, thermal and charge-coupling devices (CCD), mounted on a drone compose the embedded system being proposed for solar panels inspection. The inspection method needs more time for capturing and processing the frames and detecting the faulty panels. The system can determine the longitude and latitude of the defect position information in real-time. In this work, we investigate parallel processing for the image processing operations which reduces the processing time for the inspection systems. The results show a super-linear speedup for real-time condition monitoring in large-scale solar systems. Using the multiprocessing module in Python, we execute fault detection algorithms using streamed frames from both video cameras. The experimental results show a super-linear speedup for thermal and CCD video processing, the execution time is efficiently reduced with an average of 3.1 times and 6.3 times using 2 processes and 4 processes respectively.

Penetration detection with intention recognition for cooperatively controlled robotic needle insertion

In lumbar puncture surgeries, force and position information throughout the insertion procedure is vital for needle tip localization, because it reflects different tissue properties. Especially in pediatric cases, the changes are always insignificant for surgeons to sense the crucial feeling of loss of resistance. In this study, a robot system is developed to tackle the major clinical difficulties. Four different control algorithms with intention recognition ability are applied on a novel lumbar puncture robot system for better human–robot cooperation. Specific penetration detection based on force and position derivatives captures the feeling of loss of resistance, which is deemed crucial for needle tip location. Kinematic and actuation modeling provides a clear description of the hardware setup. The control algorithm experiment compares the human–robot cooperation performance of proposed algorithms. The experiment also dictates the clear role of designed penetration detection criteria in capturing the penetration, improving the success rate, and ensuring operational safety.

Speed Sensorless Control Based on Initial Rotor Position Detection for EESM

Electrically excited synchronous motor (EESM) is widely used in many large equipment drives because of its strong overload capacity, high efficiency, and adjustable power factor. The research and development of a high-performance EESM control system can realize the high combination of energy-saving speed regulation and green environmental protection and has a high social effect and economic value. In this paper, the signal injection method is used to obtain the initial rotor position information of EESM. Sliding Fourier transform is used to improve the initial position angle detection method based on the rotor signal injection method, and the improved method is compared with the traditional voltage integration method. Rotor high-frequency signal injection method was used to detect the rotor position information of the motor during operation, and the influence of the damping winding on the rotor signal injection method was analyzed. On the premise that the damping winding had no influence on the method, a method of obtaining the rotor position information of EESM without a speed sensor was designed. Finally, the speed sensorless regulation system using the initial rotor position detection method is simulated, which verifies the accuracy of the proposed speed sensorless control scheme.

Improved Path Planning for Indoor Patrol Robot Based on Deep Reinforcement Learning

To solve the problems of poor exploration ability and convergence speed of traditional deep reinforcement learning in the navigation task of the patrol robot under indoor specified routes, an improved deep reinforcement learning algorithm based on Pan/Tilt/Zoom(PTZ) image information was proposed in this paper. The obtained symmetric image information and target position information are taken as the input of the network, the speed of the robot is taken as the output of the next action, and the circular route with boundary is taken as the test. The improved reward and punishment function is designed to improve the convergence speed of the algorithm and optimize the path so that the robot can plan a safer path while avoiding obstacles first. Compared with Deep Q Network(DQN) algorithm, the convergence speed after improvement is shortened by about 40%, and the loss function is more stable.

Formation Control Technology of Fixed-Wing UAV Swarm Based on Distributed Ad Hoc Network

The formation control technology of the unmanned aerial vehicle (UAV) swarm is a current research hotspot, and formation switching and formation obstacle avoidance are vital technologies. Aiming at the problem of formation control of fixed-wing UAVs in distributed ad hoc networks, this paper proposed a route-based formation switching and obstacle avoidance method. First, the consistency theory was used to design the UAV swarm formation control protocol. According to the agreement, the self-organized UAV swarm could obtain the formation waypoint according to the current position information, and then follow the corresponding rules to design the waypoint to fly around and arrive at the formation waypoint at the same time to achieve formation switching. Secondly, the formation of the obstacle avoidance channel was obtained by combining the geometric method and an intelligent path search algorithm. Then, the UAV swarm was divided into multiple smaller formations to achieve the formation obstacle avoidance. Finally, the abnormal conditions during the flight were handled. The simulation results showed that the formation control technology based on distributed ad hoc network was reliable and straightforward, easy to implement, robust in versatility, and helpful to deal with the communication anomalies and flight anomalies with variable topology.

Utilization of Location-Based Services Using Augmented Reality Perception Method For Android Based School Search

Data technology has become the main means for activities in various sectors of life, one of which is location-based services for SMA/SMK schools in the Bangka Belitung area. By using a location-based service application, it is expected to be able to overcome the problem of finding the position of SMA/SMK in the Bangka Belitung area. In this study, we would like to discuss the search for SMA/SMK positions. Through the push, Location-Based Service will be displayed in the smartphone, which will help identify the presence of the school's position on the smartphone. The result of the application that is formed is an Android-based application that can recognize the presence of SMA/SMK positions. This study aims to create a location-based service application that is combined with the concept of overlaying geographic position information from augmented reality perception, which is one type of augmented reality. The overall application quality test results include 1). Based on the Functionality aspect, the percentage of the actual score got a total score of 91.8%. 2). Based on the aspect of reliability, the percentage of the actual score got a total score of 97.6%. 3). Based on the Usability aspect, the percentage of the actual score gets a total score of 92.2%. 4). Based on the aspect of efficiency, the percentage of the actual score gets a total value of 76.5%. 5). From the calculation of the total score in percent is 89.5%. So based on the total score in percent, it can be concluded that the overall application quality test results are 89.5%, so the prototypes produced in this study are included in very good criteria.

FCAU-Net for the Semantic Segmentation of Fine-Resolution Remotely Sensed Images

The semantic segmentation of fine-resolution remotely sensed images is an urgent issue in satellite image processing. Solving this problem can help overcome various obstacles in urban planning, land cover classification, and environmental protection, paving the way for scene-level landscape pattern analysis and decision making. Encoder-decoder structures based on attention mechanisms have been frequently used for fine-resolution image segmentation. In this paper, we incorporate a coordinate attention (CA) mechanism, adopt an asymmetric convolution block (ACB), and design a refinement fusion block (RFB), forming a network named the fusion coordinate and asymmetry-based U-Net (FCAU-Net). Furthermore, we propose novel convolutional neural network (CNN) architecture to fully capture long-term dependencies and fine-grained details in fine-resolution remotely sensed imagery. This approach has the following advantages: (1) the CA mechanism embeds position information into a channel attention mechanism to enhance the feature representations produced by the network while effectively capturing position information and channel relationships; (2) the ACB enhances the feature representation ability of the standard convolution layer and captures and refines the feature information in each layer of the encoder; and (3) the RFB effectively integrates low-level spatial information and high-level abstract features to eliminate background noise when extracting feature information, reduces the fitting residuals of the fused features, and improves the ability of the network to capture information flows. Extensive experiments conducted on two public datasets (ZY-3 and DeepGlobe) demonstrate the effectiveness of the FCAU-Net. The proposed FCAU-Net transcends U-Net, Attention U-Net, the pyramid scene parsing network (PSPNet), DeepLab v3+, the multistage attention residual U-Net (MAResU-Net), MACU-Net, and the Transformer U-Net (TransUNet). Specifically, the FCAU-Net achieves a 97.97% (95.05%) pixel accuracy (PA), a 98.53% (91.27%) mean PA (mPA), a 95.17% (85.54%) mean intersection over union (mIoU), and a 96.07% (90.74%) frequency-weighted IoU (FWIoU) on the ZY-3 (DeepGlobe) dataset.

Dynamic model of macro crowd merging based on abnormal pedestrian posture

Crowd merging is a complex process, and any sudden external or internal disturbance will destroy the stability of the crowd. The occurrence of abnormal behavior will affect the crowd flow process and inevitably affect the stability of the crowd flow system. The position information of the joint points is obtained through the OpenPose algorithm, and the kinematics characteristics of each node are studied. It is judged whether the number of pedestrians in the crowd and the scale of the building scene are greater than the empirical setting value based on engineering statistical data and expert experience. When the number of pedestrians is more than 2,000 and the total area of the passage is more than 2,000 square meters, the appropriate macro-dynamic model is selected. The Aw-Rascle (AR) fluid dynamics model is selected in this study. The joint point information obtained through the OpenPose is combined with the macroscopic fluid dynamics model to construct a macroscopic crowd flow dynamics model based on the pedestrian's abnormal posture.

Machine Learning on Soccer Players Position

During the last few years, sports analytics has been growing rapidly. The main usage of this discipline is the prediction of soccer match results, even if it can be applied with interesting results in different areas, such as analysis based on the player position information. In this paper, we propose an approach aimed to recognize the player position in a soccer match, predicting the specific zone in which the player is located in a specific moment. Similar objectives have never been considered yet with our best knowledge. We consider supervised machine learning techniques by considering a dataset obtained through video capturing and tracking system. The data analyzed refer to several professional soccer games captured at the Alfheim Stadium in Tromso, Norway. The approach can be used in real-time, in order to verify if a player is playing according to the guidelines of the coach. In the experimental analysis, three different types of classification have been performed, i.e., three different divisions of the field, reaching the best results with Random Tree Algorithm.

The geoPebble System: Design and Implementation of a Wireless Sensor Network of GPS-Enabled Seismic Sensors for the Study of Glaciers and Ice Sheets

The geoPebble system is a network of wirelessly interconnected seismic and GPS sensor nodes with geophysical sensing capabilities for the study of ice sheets in Antarctica and Greenland, as well as mountain glaciers. We describe our design methodology, which has enabled us to develop these state-of-the art units using commercial-off-the-shelf hardware combined with custom-designed hardware and software. Each geoPebble node is a self-contained, wirelessly connected sensor for collecting seismic activity and position information. Each node is built around a three-component seismic recorder, which includes an amplifier, filter, and 24-bit analog-to-digital converter that can sample incoming seismic signals up to 10 kHz. The timing for each node is available from GPS measurements and a local precision oscillator that is conditioned by the GPS timing pulses. In addition, we record the carrier-phase measurement of the L1 GPS signal in order to determine location at sub-decimeter accuracy (relative to other geoPebble nodes within a radius of a few kilometers). Each geoPebble includes 32 GB of solid-state storage, wireless communications capability to a central supervisory unit, and auxiliary measurements capability (including tilt from accelerometers, absolute orientation from magnetometers, and temperature). The geoPebble system has been successfully validated in the field in Antarctica and Greenland.