| 編號： | 14 |
| 期刊或研討會名稱： |
International Journal of Science and Engineering
|
| 著作名稱： | Design and Implementation of Tire Inspection System based on Deep Learning Image Recognition Technology
|
| 年度： | 2025 |
| 類別： |
期刊論文
|
| 摘要： | Currently, most tire inspection systems on the market rely on measuring tire pressure or temperature to determine whether the tires are abnormal. However, these systems cannot effectively predict tire lifespan, identify aging levels, or detect abnormal tire conditions. Neglecting the condition of ones own tires during normal driving may lead to unforeseen accidents. Therefore, this project aims to develop a tire anomaly detection system based on image recognition technology. The proposed system utilizes image recognition, machine learning, and AI algorithms to predict tire lifespan and aging levels. It distinguishes between the aging levels and abnormal conditions of tire treads. Additionally, the project includes the development of a smartphone app and a web management platform. The objectives include: 1. Instantaneous inspection through on-the-spot photo capture, allowing for a quick assessment of tire condition. 2. A mobile application featuring tire inspection, tire usage records, estimated mileage, user feedback, and other functionalities. 3. Utilization of deep learning convolutional neural network technology for recognizing various tire anomaly conditions. 4. Establishment of a cross-platform tire management web interface for use by automotive manufacturers and general users. Through experimental testing, the system demonstrates an impressive AI recognition accuracy of up to 93%. |
| 關鍵字： | machine learning, image recognition, tire inspection |
| 編號： | 12 |
| 期刊或研討會名稱： |
IEEE Transactions on Instrumentation and Measurement
|
| 著作名稱： | Enhanced PTZ Camera Dispatch Scheme for 3-D Environments Based on Deep Reinforcement Learning
|
| 年度： | 2025 |
| 類別： |
期刊論文
|
| 摘要： | In recent years, pan-tilt-zoom (PTZ) cameras with horizontal, vertical rotation, and zoom capabilities have enhanced the flexibility of fixed cameras. However, the existing research on PTZ cameras has not adequately addressed key monitoring requirements, such as viewing angle and resolution pixels per foot (PPF), often resulting in skewed perspectives or insufficient imaging quality. This study focuses on 3-D environments and aims to optimize PTZ camera dispatch for maximum target coverage while minimizing the number of cameras used. We formulate the problem as a mixed-integer linear programming problem, which is known to be NP-hard, and propose an intelligent solution using fast-prioritized experience deep Q-network (FPE-DDQN). This method leverages self-awareness and iterative learning to improve decision-making by incorporating past dispatch experiences and an innovative reward function design. FPE-DDQN prioritizes cameras based on their coverage potential and dispatch cost, significantly reducing the number of cameras needed while accelerating the adjustment of rotation angles and focal lengths. Experimental simulations demonstrate that FPE-DDQN achieves near-optimal coverage and dispatch efficiency, with the performance gap to the optimal solution being less than 0.52%–4.7%, while requiring only 1/10 to 1/100 of the computational time compared to the optimal solution. |
| 關鍵字： | Monitor resolution, pan-tilt-zoom (PTZ) camera, reinforcement learning, video surveillance, view angle |
| 編號： | 13 |
| 期刊或研討會名稱： |
IEEE Internet of Things Journal
|
| 著作名稱： | Enhanced Cell Clustering and Multicast Scheduling for Energy-Efficient 5G/B5G MBSFN Networks
|
| 年度： | 2025 |
| 類別： |
期刊論文
|
| 摘要： | The emergence of 5G and Beyond 5G (5G/B5G) networks represents a pivotal advancement in mobile wireless communication, offering peak transmission rates up to 20 Gbps. This unprecedented capacity enables a diverse range of broadband multimedia services, including IPTV and Voice/Video-over-IP applications. With the exponential growth in multimedia services, the likelihood of multiple User Equipments (UEs) within the same geographic region subscribing to identical multimedia services has surged, leading to significant inefficiencies in bandwidth utilization and radio resource consumption. To address these challenges, the Multicast Broadcast Single Frequency Network (MBSFN) mechanism has been introduced, which clusters multiple cells into a unified serving area. By employing Coordinated Multi-point Transmission (CoMP), MBSFN facilitates the synchronous delivery of multicast data across multiple UEs, effectively mitigating inter-cell interference and enhancing overall system efficiency. This approach not only reduces the data reception time for UEs but also significantly conserves their energy consumption. Nevertheless, the optimal clustering of cells and scheduling of multicast transmissions remain open research problems. In this study, we tackle the dual challenges of cell clustering and multicast scheduling with the objective of minimizing UEs’ energy consumption, measured in terms of data reception time, while ensuring the stringent Quality of Service (QoS) requirements of multicast services. To this end, we propose a two-phase framework. The first phase identifies the effective cell clusters to serve as multicast areas, balancing resource efficiency with QoS constraints. The second phase leverages dynamic programming techniques to further optimize multicast scheduling within these areas, thereby minimizing UEs’ energy. Comprehensive simulation results demonstrate that the proposed scheme significantly enhances resource utilization and reduces UEs’ energy consumption, outperforming state-of-the-art approaches. |
| 關鍵字： | Coordinated Multi-point Transmission (CoMP), Cell Clustering, Green Communications, Multicast Broadcast Single frequency Network (MBSFN), Quality of Service (QoS) |
| 編號： | 11 |
| 期刊或研討會名稱： |
IEEE Sensors Journal
|
| 著作名稱： | Enhancing Person Identification for Smart Cities: Fusion of Video Surveillance and Wearable Device Data based on Machine Learning
|
| 年度： | 2024 |
| 類別： |
期刊論文
|
| 摘要： | For smart cities, video surveillance has been widely used for security and management purposes. In video surveillance, a fundamental challenge is person identification (PID), which involves promptly tagging individuals in videos with their IDs. Using RFID and fingerprint/iris/face recognition is a possible solution. However, the identification results are highly related to environmental factors, such as line of sight, lighting conditions, and distance. Fingerprint/face recognition also has privacy concerns. In this work, we show how to achieve immediate PID through two sensor data sources: (i) human objects and their pixel locations retrieved from videos and (ii) user trajectory data retrieved from wearable devices through indoor localization. By fusing these pixel trajectories and indoor trajectories, we demonstrate an enhancing-vision capability in the sense that PID can be achieved on surveillance videos even when no clear human biological features are seen. Two types of fusion are proposed: (i) similarity-based and (ii) machine learning-based. We have developed lightweight prototyping with off-the-shelf equipment and validated our results through extensive experiments. The performance evaluation showed that our system has an accuracy of up to 92% for person identification. |
| 關鍵字： | Internet of Things (IoT), localization, machine learning, sensor fusion, video surveillance. |
| 編號： | 10 |
| 期刊或研討會名稱： |
IEEE Transactions on Intelligent Vehicles
|
| 著作名稱： | Design and Implementation of Energy-Efficient Wireless Tire Sensing System with Delay Analysis for Intelligent Vehicles
|
| 年度： | 2024 |
| 類別： |
期刊論文
|
| 摘要： | The growing prevalence of Internet of Things (IoT) technologies has led to a rise in the popularity of intelligent vehicles that incorporate a range of sensors to monitor various aspects, such as driving speed, fuel usage, distance proximity and tire anomalies. Nowadays, real-time tire sensing systems play important roles for intelligent vehicles in increasing mileage, reducing fuel consumption, improving driving safety, and reducing the potential for traffic accidents. However, the current tire sensing system drains a significant vehicle energy and lacks effective collection of sensing data, which may not guarantee the immediacy of driving safety. Thus, this paper designs an energy-efficient wireless tire sensing system (WTSS), which leverages energy-saving techniques to significantly reduce power consumption while ensuring data retrieval delays during real-time monitoring. Additionally, we mathematically analyze the worst-case transmission delay of the system to ensure the immediacy based on the collision probabilities of sensor transmissions. This system has been implemented and verified by the simulation and field trial experiments. These results show that the proposed scheme provides enhanced performance in energy efficiency up to 60.7% on average and accurately identifies the worst transmission delay. |
| 關鍵字： | Delay analysis, Internet-of-Things (IoT), energy saving, intelligent vehicles, wireless tire sensing system (WTSS). |
| 編號： | 9 |
| 期刊或研討會名稱： |
IEEE Access
|
| 著作名稱： | Discontinuous Reception Based Energy-Efficient User Association for 5G Heterogeneous Networks
|
| 年度： | 2024 |
| 類別： |
期刊論文
|
| 摘要： | The 5G Heterogeneous Network (HetNet) enables massive connections and diverse applications for User Equipments (UEs), leading to an exponential increase in data traffic. This surge results in extensive overloading at co-tier and inter-tier levels, inefficiencies in UE power usage at the cell edge, and complexities in Quality of Service (QoS) support. In this paper, we propose a scheme to address load balancing among Base Stations (BSs), while simultaneously improving power efficiency and ensuring QoS for UEs. The proposed heuristic scheme comprises three phases. The first phase optimizes the Discontinuous Reception (DRX) configuration parameters for UEs, the second phase evaluates overloading, and the third phase offloads data from overloaded BSs to other BSs. We categorize performance indicators into 1) User Performance Parameters (UPP), including DRX power saving, packet drop rate, and end-to-end delay, and 2) Network Performance Parameters (NPP), such as system throughput. To validate our scheme, we design an analytical model based on a two-dimensional continuous-time Markov chain (2D-CTMC) and a semi-Markov process. In addition, we implement a simulator to validate the scheme’s performance. The results demonstrate that the proposed scheme significantly enhances performance in terms of UPP and NPP. |
| 關鍵字： | Continuous-time Markov chain (CTMC), discontinuous reception/transmission (DRX/DTX), heterogeneous network (HetNet), load balancing, quality of service (QoS), semi-Markov process. |
| 編號： | 8 |
| 期刊或研討會名稱： |
IEEE Sensors Letters
|
| 著作名稱： | Dynamic Sleep Scheduling for Wireless TP Sensor Transmissions Based on Reinforcement Learning
|
| 年度： | 2023 |
| 類別： |
期刊論文
|
| 摘要： | Tire Pressure (TP) sensors are an essential component of the Tire Pressure Monitoring System (TPMS) for smart vehicles. Each TP sensor is mounted on a wheel for real-time detecting and monitoring abnormal tire conditions to ensure driving safety. In the TPMS system, multiple TP sensors transmit sensing data randomly at a regular period to the TPMS receiver with a special communication module of 433 MHz. Traditionally, the TPMS receiver adopts a power saving mechanism to perform wake-up and sleep operation within a fixed cycle length for sensing data collection to save energy. However, the transmission delay may become worse once the number of TP sensors increases. Therefore, this letter tries to consider such problem and develops a dynamic scheduling approach for the TPMS by balancing the energy efficiency while ensuring data transmission delay. Specifically, the proposed approach exploits a reinforcement learning approach to dynamically control the wake-up and sleep periods of the TPMS receiver based on the designed reward function. Simulation results show that the proposed approach can significantly decrease data transmission delay by 21.94%-64.13% in average and improve energy efficiency by 5.29%-39.57% in average, as compared with the existing schemes. |
| 關鍵字： | Sensor Systems, dynamic scheduling, energy efficiency, reinforcement learning, tire pressure (TP) sensor transmission, transmission delay. |
| 編號： | 7 |
| 期刊或研討會名稱： |
International Computer Symposium
台灣
|
| 著作名稱： | Design and Analysis for Wireless Tire Pressure Sensing System
|
| 年度： | 2023 |
| 類別： |
會議論文
|
| 摘要： | With the expansion of Internet of Things (IoT) technologies, intelligent vehicles are now in increasing public demand, mostly equipped with sensors for measuring various situations, such as distance proximity, driving speed, fuel consumption, and tire anomaly. Nowadays, tire pressure and temperature play an important role in reducing fuel consumption, increasing mileage, improving driving safety, and reducing potential traffic accidents. In this paper, we investigate how to collect the tire data based on the designed sensors and then develop an energy-efficient system, called Wireless Tire Pressure Sensing System (WTPSS), which leverages the energy-saving technique by applying wakeup-sleep operation, while ensuring the data retrieval delay when monitoring. In addition, the worst receiving delay is also discussed and analyzed theoretically. In the simulation, our system has verified that it can have better performance in terms of energy saving. |
| 關鍵字： | Tire pressure sensing system, Worst-delay, Sensor data transmission, Wireless transmission |
| 編號： | 4 |
| 期刊或研討會名稱： |
International Journal of Science and Engineering
|
| 著作名稱： | Smart Infusion System based on Physiological Sensing and Machine Learning Technology
|
| 年度： | 2022 |
| 類別： |
期刊論文
|
| 摘要： | In response to the increasing nurse-to-patient ratio and the need to improve the quality of patient care, this study focuses on intravenous (IV) infusion, a common medical procedure. We propose the design of a system that integrates the Internet of Things (IoT) and machine learning technologies to provide the following three core functions: (1) a web interface for real-time display of patients’ physiological data, (2) a device for measuring and controlling infusion rates, and (3) a machine learning–based module that suggests optimal infusion speeds and alerts caregivers to abnormal patient conditions.
For the front-end hardware, this study adopts Arduino as the IoT communication platform for the sensors, enabling wireless transmission of collected data and reception of control commands from the back end. On the back end, Node.js is used to build both the IoT and web servers, allowing real-time data visualization and remote control. To further enhance the interpretation of patients physiological conditions, the system incorporates machine learning to analyze the incoming data. This assists nursing staff in delivering faster and more accurate treatment, thereby achieving precise infusion care through technological support. |
| 關鍵字： | physiological sensing, machine learning, wireless networking, infusion control, Internet of Things |
| 編號： | 5 |
| 期刊或研討會名稱： |
International Journal of Science and Engineering
|
| 著作名稱： | The Eye of Healthcare: Visualization of Personal Physiological Condition
|
| 年度： | 2022 |
| 類別： |
期刊論文
|
| 摘要： | In recent years, the average age of patients with cardiovascular diseases has been decreasing, while myocardial infarction remains the leading cause of mortality. Traditional diagnostic methods for heart disease do not offer real-time and effective monitoring capabilities, preventing timely medical intervention. This study proposes a real-time AR-based visualization system for monitoring the physiological status of individuals within a given environment. The system enables feature extraction and ECG signal analysis directly on wearable devices. To address privacy concerns associated with RGB cameras, the proposed approach integrates RPLiDAR with inertial sensing wearables to perform identity recognition and tracking. Through augmented reality visualization, the physiological data of each monitored individual is displayed in real time on the surveillance interface. Physiological data is continuously recorded and analyzed using feature extraction techniques. By leveraging machine learning models to detect myocardial ischemia and applying sensor-fusion-based identity recognition techniques, the system can immediately alert caregivers to abnormal physiological conditions, enabling timely preventive measures. Experimental results demonstrate that the algorithms developed in this system outperform existing methods in terms of accuracy and efficiency. |
| 關鍵字： | Myocardial Ischemia, LiDAR, Inertial Sensors, Internet of Things, Data Fusion |
| 編號： | 3 |
| 期刊或研討會名稱： |
The 26th Mobile Computing Workshop
台灣
|
| 著作名稱： | Tire Anomaly Detection System Based on Machine Learning and IoT Sensing Technology
|
| 年度： | 2022 |
| 類別： |
會議論文
|
| 摘要： | 傳統市售的胎壓偵測器僅能透過測量胎壓、胎溫來判斷輪胎狀態，無法即時檢測輪胎的異常狀況，如：輪框變形、胎面異常剝離、輪胎附著異物、異常抖動及打滑。一旦駕駛疏於注意，將造成嚴重的車禍。因此，我們研究設計一套智慧輪胎異常偵測系統，透過量測輪胎運行間的慣性感測資訊，包含：多軸加速度、角速度、磁力計，再結合機器學習之辨識技術，以動態檢測出輪胎異常狀況，並實作設計下述物聯網平台元件：1. 偵測輪胎慣性感測數值之硬體裝置, 2. 具即時監控、異常提醒、歷史紀錄與統計等功能之手機App, 3. 基於機器學習技術辨識各式輪胎異常狀況, 4. 建立跨平台網頁介面供停車場管理車輛、資料蒐集與大數據分析。經實驗測試結果，本系統的AI判讀準確率高達95%。 |
| 關鍵字： | 物聯網、機器學習、智慧終端裝置、輪胎異常、無線網路 |
| 編號： | 6 |
| 期刊或研討會名稱： |
Sensors
|
| 著作名稱： | Applying Image Recognition and Tracking Methods for Fish Physiology Detection Based on a Visual Sensor
|
| 年度： | 2022 |
| 類別： |
期刊論文
|
| 摘要： | The proportion of pet keeping has increased significantly. According to the survey results of Business Next, the proportion of Taiwan families keeping pets was 70% in 2020. Among them, the total number of fish pets was close to 33% of the overall pet proportion. Therefore, aquarium pets have become indispensable companions for families. At present, many studies have discussed intelligent aquarium systems. Through image recognition based on visual sensors, we may be able to detect and interpret the physiological status of the fish according to their physiological appearance. In this way, it can help to notify the owner as soon as possible to treat the fish or isolate them individually, so as to avoid the spread of infection. However, most aquarium pets are kept in groups. Traditional image recognition technologies often fail to recognize each fish’s physiological states precisely because of fish swimming behaviors, such as grouping swimming, shading with each other, flipping over, and so on. In view of this, this paper tries to address such problems and then proposes a practical scheme, which includes three phases. Specifically, the first phase tries to enhance the image recognition model for small features based on the prioritizing rules, thus improving the instant recognition capability. Then, the second phase exploits a designed fish-ID tracking mechanism and analyzes the physiological state of the same fish-ID through coherent frames, which can avoid temporal misidentification. Finally, the third phase leverages a fish-ID correction mechanism, which can detect and correct their IDs periodically and dynamically to avoid tracking confusion, and thus potentially improve the recognition accuracy. According to the experiment results, it was verified that our scheme has better recognition performance. The best accuracy and correctness ratio can reach up to 94.9% and 92.67%, which are improved at least 8.41% and 26.95%, respectively, as compared with the existing schemes. |
| 關鍵字： | image recognition; object tracking; correction mechanism; Internet of Things |
| 編號： | 2 |
| 期刊或研討會名稱： |
Sensors
|
| 著作名稱： | Applying Machine Learning Technologies Based on Historical Activity Features for Multi-Resident Activity Recognition
|
| 年度： | 2021 |
| 類別： |
期刊論文
|
| 摘要： | Due to the aging population, home care for the elderly has become very important. Currently, there are many studies focusing on the deployment of various sensors in the house to recognize the home activities of the elderly, especially for the elderly living alone. Through these, we can detect the home situation of the single person and ensure his/her living safety. However, the living environment of the elderly includes, not only the person living alone, but also multiple people living together. By applying the traditional methods for a multi-resident environment, the “individual” activities of each person could not be accurately identified. This resulted in an inability to distinguish which person was involved in what activities, and thus, failed to provide personal care. Therefore, this research tries to investigate how to recognize home activities in multi-resident living environments, in order to accurately distinguish the association between residents and home activities. Specifically, we propose to use the special characteristics of historical activity of residents in a multi-person environment, including activity interaction, activity frequency, activity period length, and residential behaviors, and then apply a suite of machine learning methods to train and test. Five traditional models of supervised learning and two deep learning methods are explored to tackle this problem. Through the experiments with real datasets, the proposed methods were found to achieve higher precision, recall and accuracy with less training time. The best accuracy can reach up to 91% and 95%, by J48DT, and LSTM, respectively, in different living environments. |
| 關鍵字： | ambient assisted living; multi-person activity recognition; machine learning; deep learning |
| 編號： | 1 |
| 期刊或研討會名稱： |
Sensors
|
| 著作名稱： | Dynamic Set Planning for Coordinated Multi-Point in B4G/5G Networks
|
| 年度： | 2021 |
| 類別： |
期刊論文
|
| 摘要： | Coordinated Multi-Point (CoMP) is an important technique in B4G/5G networks. With CoMP, multiple base stations can be clustered to compose a cooperating set to improve system throughput, especially for the users in cell edges. Existed studies have discussed how to mitigate overloading scenarios and enhance system throughput with CoMP statically. However, static cooperation fixes the set size and neglects the fast-changing of B4G/5G networks. Thus, this paper provides a full study of off-peak hours and overloading scenarios. During off-peak hours, we propose to reduce BSs’ transmission power and use the free radio resource to save energy while guaranteeing users’ QoS. In addition, if large-scale activities happen with crowds gathering or in peak hours, we dynamically compose the cooperating set based on instant traffic requests to adjust base stations’ BSs’ transmission power; thus, the system will efficiently offload the traffic to the member cells which have available radio resources in the cooperating set. Experimental results show that the proposed schemes enhance system throughput, radio resource utilization, and energy efficiency, compared to other existing schemes. |
| 關鍵字： | coordinated multi-point (CoMP); cooperating set; dynamic cell selection (DCS); energy efficiency; beyond fourth-generation/fifth-generation (B4G/5G); resource allocation; soft frequency reuse (SFR) |