Background: A stroke often bequeaths surviving patients with impaired neuromusculoskeletal systems subjecting them to increased risk of injury (e.g., due to falls) even during activities of daily living. The risk of injuries to such individuals can be related to alterations in their movement. Using inertial sensors to record the digital biomarkers during turning could reveal the relevant turning alterations. Objectives: In this study, movement alterations in stroke survivors (SS) were studied and compared to healthy individuals (HI) in the entire turning task due to its requirement of synergistic application of multiple bodily systems. Methods: The motion of 28 participants (14 SS, 14 HI) during turning was captured using a set of four Inertial Measurement Units, placed on their sternum, sacrum, and both shanks. The motion signals were segmented using the temporal and spatial segmentation of the data from the leading and trailing shanks. Several kinematic parameters, including the range of motion and angular velocity of the four body segments, turning time, the number of cycles involved in the turning task, and portion of the stance phase while turning, were extracted for each participant. Results: The results of temporal processing of the data and comparison between the SS and HI showed that SS had more cycles involved in turning, turn duration, stance phase, range of motion in flexion–extension, and lateral bending for sternum and sacrum (p-value < 0.035). However, HI exhibited larger angular velocity in flexion–extension for all four segments. The results of the spatial processing, in agreement with the prior method, showed no difference between the range of motion in flexion–extension of both shanks (p-value > 0.08). However, it revealed that the angular velocity of the shanks of leading and trailing legs in the direction of turn was more extensive in the HI (p-value < 0.01). Conclusions: The changes in upper/lower body segments of SS could be adequately identified and quantified by IMU sensors. The identified kinematic changes in SS, such as the lower flexion–extension angular velocity of the four body segments and larger lateral bending range of motion in sternum and sacrum compared to HI in turning, could be due to the lack of proper core stability and effect of turning on vestibular system of the participants. This research could facilitate the development of a targeted and efficient rehabilitation program focusing on the affected aspects of turning movement for the stroke community. Full article

Mobile construction machineries are accident-prone on a dynamic construction site, as the site environment is constantly changing and continuous safety monitoring by human beings is impossible. These accidents usually happen in the form of machinery overturning or collapsing into risk areas, including the foundation pit, slopes, or soft soil area. Therefore, preventing mobile construction machineries from entering risk areas is the key. However, currently, there is a lack of practical safety management techniques to achieve this. Utilizing a wireless sensor device to collect the location information of mobile construction machineries, this research develops a safety warning algorithm to prevent the machineries moving into risk area and reduces onsite overturning or collapsing accidents. A modified axis aligned bounding box method is proposed according to the movement patterns of mobile construction machineries, and the warning algorithm is developed based on the onsite safety management regulations. The algorithm is validated in a real case simulation when machinery enters the warning zone. The simulation results showed that the overall algorithm combining the location sensing technology and the modified bounding box method could detect risk and give warnings in a timely manner. This algorithm can be implemented for the safety monitoring of mobile construction machineries in daily onsite management. Full article

Gamification is known to enhance users’ participation in education and research projects that follow the citizen science paradigm. The Cosmic Ray Extremely Distributed Observatory (CREDO) experiment is designed for the large-scale study of various radiation forms that continuously reach the Earth from space, collectively known as cosmic rays. The CREDO Detector app relies on a network of involved users and is now working worldwide across phones and other CMOS sensor-equipped devices. To broaden the user base and activate current users, CREDO extensively uses the gamification solutions like the periodical Particle Hunters Competition. However, the adverse effect of gamification is that the number of artefacts, i.e., signals unrelated to cosmic ray detection or openly related to cheating, substantially increases. To tag the artefacts appearing in the CREDO database we propose the method based on machine learning. The approach involves training the Convolutional Neural Network (CNN) to recognise the morphological difference between signals and artefacts. As a result we obtain the CNN-based trigger which is able to mimic the signal vs. artefact assignments of human annotators as closely as possible. To enhance the method, the input image signal is adaptively thresholded and then transformed using Daubechies wavelets. In this exploratory study, we use wavelet transforms to amplify distinctive image features. As a result, we obtain a very good recognition ratio of almost 99% for both signal and artefacts. The proposed solution allows eliminating the manual supervision of the competition process. Full article

In this paper, we propose a novel approach that enables simultaneous localization, mapping (SLAM) and objects recognition using visual sensors data in open environments that is capable to work on sparse data point clouds. In the proposed algorithm the ORB-SLAM uses the current and previous monocular visual sensors video frame to determine observer position and to determine a cloud of points that represent objects in the environment, while the deep neural network uses the current frame to detect and recognize objects (OR). In the next step, the sparse point cloud returned from the SLAM algorithm is compared with the area recognized by the OR network. Because each point from the 3D map has its counterpart in the current frame, therefore the filtration of points matching the area recognized by the OR algorithm is performed. The clustering algorithm determines areas in which points are densely distributed in order to detect spatial positions of objects detected by OR. Then by using principal component analysis (PCA)—based heuristic we estimate bounding boxes of detected objects. The image processing pipeline that uses sparse point clouds generated by SLAM in order to determine positions of objects recognized by deep neural network and mentioned PCA heuristic are main novelties of our solution. In contrary to state-of-the-art approaches, our algorithm does not require any additional calculations like generation of dense point clouds for objects positioning, which highly simplifies the task. We have evaluated our research on large benchmark dataset using various state-of-the-art OR architectures (YOLO, MobileNet, RetinaNet) and clustering algorithms (DBSCAN and OPTICS) obtaining promising results. Both our source codes and evaluation data sets are available for download, so our results can be easily reproduced. Full article

The design and implementation of an electronic system that involves head movements to operate a prototype that can simulate future movements of a wheelchair was developed here. The controller design collects head-movements data through a MEMS sensor-based motion capture system. The research was divided into four stages: First, the instrumentation of the system using hardware and software; second, the mathematical modeling using the theory of dynamic systems; third, the automatic control of position, speed, and orientation with constant and variable speed; finally, system verification using both an electronic controller test protocol and user experience. The system involved a graphical interface for the user to interact with it by executing all the controllers in real time. Through the System Usability Scale (SUS), a score of 78 out of 100 points was obtained from the qualification of 10 users who validated the system, giving a connotation of “very good”. Users accepted the system with the recommendation to improve safety by using laser sensors instead of ultrasonic range modules to enhance obstacle detection. Full article

Human pose estimation and tracking in real-time from multi-sensor systems is essential for many applications. Combining multiple heterogeneous sensors increases opportunities to improve human motion tracking. Using only a single sensor type, e.g., inertial sensors, human pose estimation accuracy is affected by sensor drift over longer periods. This paper proposes a human motion tracking system using lidar and inertial sensors to estimate 3D human pose in real-time. Human motion tracking includes human detection and estimation of height, skeletal parameters, position, and orientation by fusing lidar and inertial sensor data. Finally, the estimated data are reconstructed on a virtual 3D avatar. The proposed human pose tracking system was developed using open-source platform APIs. Experimental results verified the proposed human position tracking accuracy in real-time and were in good agreement with current multi-sensor systems. Full article

Autonomous systems need to localize and track surrounding objects in 3D space for safe motion planning. As a result, 3D multi-object tracking (MOT) plays a vital role in autonomous navigation. Most MOT methods use a tracking-by-detection pipeline, which includes both the object detection and data association tasks. However, many approaches detect objects in 2D RGB sequences for tracking, which lacks reliability when localizing objects in 3D space. Furthermore, it is still challenging to learn discriminative features for temporally consistent detection in different frames, and the affinity matrix is typically learned from independent object features without considering the feature interaction between detected objects in the different frames. To settle these problems, we first employ a joint feature extractor to fuse the appearance feature and the motion feature captured from 2D RGB images and 3D point clouds, and then we propose a novel convolutional operation, named RelationConv, to better exploit the correlation between each pair of objects in the adjacent frames and learn a deep affinity matrix for further data association. We finally provide extensive evaluation to reveal that our proposed model achieves state-of-the-art performance on the KITTI tracking benchmark. Full article