Existing wearable systems that use G-sensors to identify daily activities have been widely applied for medical, sports and military applications, while body temperature as an obvious physical characteristic that has rarely been considered in the system design and relative applications of HAR. In the context of the normalization of COVID-19, the prevention and control of the epidemic has become a top priority. Temperature monitoring plays an important role in the preliminary screening of the population for fever. Therefore, this paper proposes a wearable device embedded with inertial and temperature sensors that is used to apply human behavior recognition (HAR) to body surface temperature detection for body temperature monitoring and adjustment by evaluating recognition algorithms. The sensing system consists of an STM 32-based microcontroller, a 6-axis (accelerometer and gyroscope) sensor, and a temperature sensor to capture the original data from 10 individual participants under 4 different daily activity scenarios. Then, the collected raw data are pre-processed by signal standardization, data stacking and resampling. For HAR, several machine learning (ML) and deep learning (DL) algorithms are implemented to classify the activities. To compare the performance of different classifiers on the seven-dimensional dataset with temperature sensing signals, evaluation metrics and the algorithm running time are considered, and random forest (RF) is found to be the best-performing classifier with 88.78% recognition accuracy, which is higher than the case of the absence of temperature data (<78%). In addition, the experimental results show that participants’ body surface temperature in dynamic activities was lower compared to sitting, which can be associated with the possible missing fever population due to temperature deviations in COVID-19 prevention. According to different individual activities, epidemic prevention workers are supposed to infer the corresponding standard normal body temperature of a patient by referring to the specific values of the mean expectation and variance in the normal distribution curve provided in this paper.

Activity recognition has received considerable attention in many research fields, such as industrial and healthcare fields. However, many researches about activity recognition have focused on static activities and dynamic activities in current literature, while, the transitional activities, such as stand-to-sit and sit-to-stand, are more difficult to recognize than both of them. Consider that it may be important in real applications. Thus, a novel framework is proposed in this paper to recognize static activities, dynamic activities, and transitional activities by utilizing stacked denoising autoencoders (SDAE), which is able to extract features automatically as a deep learning model rather than utilize manual features extracted by conventional machine learning methods. Moreover, the resampling technique (random oversampling) is used to improve problem of unbalanced samples due to relatively short duration characteristic of transitional activity. The experiment protocol is designed to collect twelve daily activities (three types) by using wearable sensors from 10 adults in smart lab of Ulster University, the experiment results show the significant performance on transitional activity recognition and achieve the overall accuracy of 94.88% on three types of activities. The results obtained by comparing with other methods and performances on other three public datasets verify the feasibility and priority of our framework. This paper also explores the effect of multiple sensors (accelerometer and gyroscope) to determine the optimal combination for activity recognition.

Integrated Sensing and Communication (ISAC) is one of the most promising technical directions in B5G/6G era. Nevertheless, it is challenging for network operators to provide proper Human Centric Services (HCS), appropriately with the explosive growth of mobile terminals in constantly changing scenarios. Scenario sensing is the foundation to provide users with intelligent context-aware services, while the combination of machine learning (ML) and automatic data collection enables network operators to become more flexible in decision-making and planning. In this paper, a machine learning-based integrated indoor/outdoor (IO) sensing and positioning framework is proposed to improve the potential of ML in the context of cellular network service management. Firstly, large amounts of measurement reports (MRs) are collected over Layer 3 at user equipment (UE) and evolved NodeBs (eNBs) in urban areas, including both indoor and outdoor samples to simulate the Minimization of Drive Test (MDT). Then, a random forest-based IO classifier is implemented to sense the mobile scenarios and filter the positioning fingerprints. Subsequently, the Weighted K-nearest Neighbor (WKNN)-based Enhanced Cell ID (ECID) merging with MRs method is used for cellular positioning. The performance evaluation has shown that the positioning error of the MRs after denoising is effectively reduced compared to the conventional fingerprint-based positioning. Specifically, under the 67% standard defined by Federal Communications Commission (FCC), the resulting positioning error is about 4% lower than the case without preprocessing. In addition, comparative experiments are conducted to discuss the impact of different K values at different data sizes of fingerprints. Moreover, the further data analysis shows MRs collected in the mild indoor scenarios have a positive effect on the overall positioning error, which indicates refined contextualizations contribute to the cellular positioning for HCS upgrade.

Human activity recognition (HAR) has received more and more attention, which is able to play an important role in many fields, such as healthcare and intelligent home. Thus, we have discussed an application of activity recognition in the healthcare field in this paper. Essential tremor (ET) is a common neurological disorder that can make people with this disease rise involuntary tremor. Nowadays, the disease is easy to be misdiagnosed as other diseases. We have combined the essential tremor and activity recognition to recognize ET patients’ activities and evaluate the degree of ET for providing an auxiliary analysis toward disease diagnosis by utilizing stacked denoising autoencoder (SDAE) model. Meanwhile, it is difficult for model to learn enough useful features due to the small behavior dataset from ET patients. Thus, resampling techniques are proposed to alleviate small sample size and imbalanced samples problems. In our experiment, 20 patients with ET and 5 healthy people have been chosen to collect their acceleration data for activity recognition. The experimental results show the significant result on ET patients activity recognition and the SDAE model has achieved an overall accuracy of 93.33%. What’s more, this model is also used to evaluate the degree of ET and has achieved the accuracy of 95.74%. According to a set of experiments, the model we used is able to acquire significant performance on ET patients activity recognition and degree of tremor assessment.

One of the main topics of Smart Home (SH) research is the recognition of activities performed by its inhabitants, which is considered to be one of the bases to foster new technological solutions inside the home, including services to prolong independent living of the elderly. However, current activity recognition proposals still find problems when considering all the different types of activities that can be performed at home, namely static, dynamic, and transitional activities. In this paper, we consider recognition of transitional activities, which is often ignored in most studies. In addition, we propose a novel dynamic segmentation method based on change points in data stream and construct an ensemble of heterogeneous classifiers to recognize twelve activities (of all types). The experiment is conducted on the dataset collected over ten hours by a wearable accelerometer placed on the person’s wrist. The base classifiers selected to form this ensemble are support vector machine (SVM), decision tree (DT) and k-nearest neighbors (KNN). As a result, the proposed approach has achieved an overall classification accuracy equal to 96.87% with 10-fold cross-validation. Moreover, all activity types considered have been similarly well identified.

With the continuous expansion of traffic volume in China, pavement deterioration recognition is a key technique to reflect the real state of the roads and to provide the basis for decision of the maintenance plan. In this paper, a new pavement defect detection system based on Image Processing was developed which utilized Unmanned Aerial Vehicle (UAV) to avoid blocking the traffic flow and to enlarge monitoring range. It was divided into two subsystems: UAV intelligent control modulus and defect detection modulus. For the former, special programs were designed to achieve the functions such as automatically taking off and landing, automatically trajectory control, and camera and gimbal control. For the latter, based on the comparative analysis of the basic principles and multiple image groups experiments of normal edge detection algorithms, a combined method was proposed. This proposed method can not only automatically provide the high threshold that Canny algorithm needs to input manually, but also solve the lack of edge detection ability in OTSU segmentation algorithm. The efficiency and accuracy of the image processing is verified through the experiments. It is also showed that the weak edges in the image can be better retained by taking 1/3 to 1/2 of the high threshold as a low threshold in this combined method. According to both qualitative and quantitative analysis, it can be concluded that, among the normal algorithms taken into study, the proposed combination algorithm performs best in terms of recognition accuracy of crack types and crack geometry features extraction.