Search

Hyperspectral anomaly detection has attracted extensive interests for its wide use in military and civilian fields, and three main categories of detection methods have been developed successively over past few decades, including statistical model-based, representation-based, and deep-learning-based methods. Most of these algorithms are essentially trying to construct proper background profiles, which describe the characteristics of background and then identify the pixels that do not conform to the profiles as anomalies. Apparently, the crucial issue is how to build an accurate background profile; however, the background profiles constructed by existing methods are not accurate enough. In this article, a novel and universal background purification framework with extended morphological attribute profiles is proposed. It explores the spatial characteristic of image and removes suspect anomaly pixels from the image to obtain a purified background. Moreover, three detectors with this framework covering different categories are also developed. The experiments implemented on four real hyperspectral images demonstrate that the background purification framework is effective, universal, and suitable. Furthermore, compared with other popular algorithms, the detectors with the framework perform well in terms of accuracy and efficiency.

Multibeam (MB) radiations of antenna arrays need effective beamforming networks (BFNs). This paper presents a fisheye lens-based BFN (FL-BFN) by parallel-plate sandwiched substrates. It implements the beam and antenna excitation ports on the half-circular circumference and straight-line boundaries. The rule to determine the positions of antenna elements is shown for easy design. The performance is validated by using it to excite a Vivaldi antenna array. Measurements over a prototype were performed to validate its feasibility with good agreements.

As a family member of the Low Power Wide Area Networks (LPWAN), LoRa provides resistance to interference and multipath fading by the Chirp Spread Spectrum (CSS) technique. It improves its performance by adjusting transmission parameters to adapt to different circumstances. This paper presents LoRa transmission parameter evaluation under different circumstances, with and without building obstacle blocking or distorting the wireless signals. The parameters include Spreading Factors (SF), Bandwidth (BW), and Coding Rate (CR). We tested different parameter combinations. Yield tests of 72 different combinations on transmission parameters are carried out to evaluate the performance under each circumstance.

Image segmentation is a basic and important strategy to extract the region of interesting (ROI). However the ROI of retinal fundus image, such as the foreground or optic disc, often could not be generated correctly, due to poor lighting environments, physiological features or improper imaging equipment. Hence, a novel fundus image segmentation method is proposed in this paper. Briefly speaking, considering each line of fundus image as a time series signal, we adopt a step time series signal polynomial curve fitting methodology to smooth the noise and to exclude all features in the fundus image without any loss of the target margin. After fitting the fundus image, a refinement tactic with Kmeans algorithm is used to obtain the final segmentation results. This unsupervised segmentation approach is a completely robust method, because all break points in the signal are detected automatically by the Buishand Range (BR) homogeneity breakpoint detection algorithm. To demonstrate the experimental performance of our new proposal, the test results of foreground segmentation have been compared with their ground truth on DRIVE and STARE datasets, and yield an encouraging performance. Based on these two public datasets and our private database which have been annotated by the medical experts, the experimental outputs of optic disk segmentation achieve the average overlap area ratio of 92.88%, outperforming existing unsupervised methods in this field.

TV networks are no longer just closed networks. They are increasingly carrying Internet services, integrating and interoperating with home IoT and the Internet. In addition, client devices are becoming intelligent. At the same time, they are facing more security risks. Security incidents such as attacks on TV systems are commonplace, and there are many incidents that cause negative effects. The security protection of TV networks mainly adopts security protection schemes similar to other networks, such as constructing a security perimeter; there are few security researches specifically carried out for client-side devices. This paper focuses on the mainstream architecture of the integration of HFC TV network and the Internet, and conducts a comprehensive security test and analysis for client-side devices including EOC cable bridge gateways and smart TV Set-Top-BoX. Results show that the TV network client devices have severe vulnerabilities such as command injection and system debugging interfaces. Attackers can obtain the system control of TV clients without authorization. In response to the results, we put forward systematic suggestions on the client security protection of smart TV networks in current days.

This research-to-practice WIP (Work In Progress) presents the design and assessment of online Model-Based Systems Engineering (MBSE) modules for practicing engineers using social constructivism as a theoretical framework. Despite many advantages of MBSE, experts in this field are still scarce in the current engineering workforce. To address this need, an online module that will be deployed in the summer of 2020 targets practicing engineers as its learners to equip them with MBSE-related knowledge and skills. In industry, teams working on MBSE-related projects usually collaborate across multidisciplinary units. Therefore, social interaction plays an integral role in MBSE training programs. To understand how group interaction could foster learning in online engineering modules, we apply social constructivism as a theoretical framework to engage learners in meaningful interactions and facilitate the acquisition and application of knowledge. The modules will utilize deep-level, student-centered, small-group discussions, and peer review between student groups as forms of social learning in authentic engineering assignments. The assessment of the modules will focus on the effectiveness of social learning in promoting the mastery and application of content knowledge.

In this paper, point cloud and image feature distribution algorithms are adopted. First, we use the depth information of the LiDAR point cloud to perform object segmentation. Secondly, we use the result of object segmentation to find the ROI (region of interest) on the color image and set it as the input of the R-YOLO neural network for object detection. Finally, our algorithm will simultaneously obtain LiDAR point cloud and image recognition results in real time. In addition, we will use a cell-based design flow to implement high-speed convolution chips to improve system performance.

The study aimed to develop an escape room educational game for learning cardiopulmonary resuscitation (CPR) procedure. To evaluate the game, thirty-eight junior high students from northern Taiwan participated in a learning activity using the game. The results revealed that learners this game enhanced students’ learning achievement of CPR procedure. Meanwhile, learners generally enjoyed playing the game in terms of their flow state, and perceived the game as useful and easy to use.

The study aimed to design a real escape gamification teaching activity for elementary school social studies education. The teaching objectives is to guide students to learn and understand the culture and history of their hometown. The teaching activity uses slides and learning sheets to present situation with challenging tasks, and offers clues and opportunity for learners to participate problem-solving process in the real world. A total of 25 elementary school students were recruited to participate in the current research. The students were required to search information and answer the questions on the Internet via mobile devices. The results revealed that this learning activity enhanced students' learning performance, flow state, and game acceptance in social education.

In terms of 5G mobile broadband technology, the application of millimeter wave is one of the key components, and phased array antenna has been an important way to realize high gain. Therefore, the suppression of the side-lobes of array antenna with millimeter wave frequency has become a major issue in antenna design. This paper focuses on the design of array antennas, dividing into two parts. (1) Analysis of array antenna design which reduces sidelobes: We first propose a design having a resistive plate placed at the array antenna aperture, so that the main beam can radiate smoothly without affecting the beam scanning. In terms of reducing side lobes, the electromagnetic wave can be attenuated by the resistive plate we design. We also design a set of resistive horn edge and rolled edge to be placed along the edges of array antenna to reduce the edge diffraction effect. (2) The development of using numerical genetic algorithm (GA) on the suppression of side-lobes of array antenna: This paper develops a radiation optimization algorithm for the resistive plate placed on the array antenna implemented in a practical numerical electromagnetic method as a design tool for developing 5G radio frequency array antennas.

Autonomous vehicles and robot are more and more popular, and it is important to sense the environment for autonomous vehicles and robot. LiDAR is a very powerful device, it can capture the distance and contour of object precisely and quickly. And the scanning range is very wide. we use 3D LiDAR and neural network to detection the environment. We get the data packet from LiDAR, and decode the data packet to 3D point cloud coordinate. Then extract the drivable road by neural network and mark the obstacle.

Nowadays, the development of automatic driving technology is well-improved the object segmentation with the point cloud data, furthermore will build up the environment model with the color image. However, the analysis of color image will still face lots of bottlenecks. This paper proposed using a generative adversarial network to achieve intrinsic image decomposition faster and then applying a new interpolation algorithm to enhance the shadow layer to increase the contrast of color image. The results show that this method improved the accuracy and immediacy of the image enhancement, go a step further, improving the accuracy and immediacy of the environment reconstruction.

In this paper, we proposed a band-stop silicon photonic filter based on the highly tunable silicon on insulator (SOI) split micro-ring resonators. The band-stop filter is made up of a silicon bus waveguide coupled to SOI split micro-ring resonators. The band-stop of the proposed silicon photonic waveguide structures can be easily manipulated by adjusting the geometrical parameters of the split micro-ring resonators. The numerical results are obtained by a finite-difference time-domain (FDTD) method. The proposed resonator would be a potential key component in the application of integrated optic circuits.

This paper presents a strategy to measure the radiation from an electrically large phased array of antennas with high-power excitations. The motivation arises from the insufficient power supply and safety requirement in the measurement system. We propose the decomposition of the original array into mathematically overlapped arrays each with locally emphasized excitations to fulfill the safety requirement of radiation inside a chamber. The radiations of these decomposed arrays may also serve as the field basis functions to calibrate the radiation patterns of the original array with sufficiently reduced time. Both numerical simulations and experimental measurements are shown to demonstrate the validity.

The Internet of Underwater Things (IoUT) is a novel class of Internet of Things (IoT) that enables practical applications toward developing smart cities. Underwater Wireless Sensor Network (UWSN) shows great potential for the future but also poses new challenges for IoUT. In addition, routing protocol will play a major role in handling how to forward data among the substantial number of things“. In this paper, we call attention to the challenges for IoUT, provide an exhaustive investigation of cutting-edge routing protocol and find the connection between the challenges and routing protocols for IoUT.

This paper presents a hierarchical MCU (micro-control unit) architecture to reduce the number of wires required to control the RF devices of large phased array of antennas. In addition to cost reduction, the decrease of control wires may also reduce the electromagnetic interferences to the antenna radiation that are caused by the presence of wires. In particular, Genetic algorithm (GA) was also implemented to find the parameters in binary codes to reduce the phase discrepancy. Numerical examples are presented to demonstrate the feasibility.

A hierarchical beamforming network (BFN) architecture to feed phased arrays of antennas for directional beam radiation is developed by using a subarray modularization technique. Instead of feeding the antenna elements by using individual sets of RF devices in a one-to-one correspondence, the RF devices are sub-grouped for modularization with each module formed by very few elements to form multi-level BFNs, whose numbers of elements are the integer factors of array size. Each module consists of identical RF structures for the minimization of manufacture and implementation cost. In this paper, examination of the discrepancy in the radiation characteristics is investigated to demonstrate the feasibility and validate the technique.

The Internet of Underwater Things (IoUT) is a novel class of Internet of Things (IoT), and one of the emerging technologies toward smart city. To support the concept of IoUT, Underwater Wireless Sensor Networks (UWSNs) have emerged as a promising system. In this paper, we survey the potential IoUT applications, point out the challenges and differences between UWSNs and traditional sensor networks, and investigate the channel models. We validated the models by simulations.

The technique of 3D scene reconstruction from 2D scene image can satisfy the demands of 3D display market, and it is the main research contents in the field of image processing. For the 3D reconstruction, the extraction of depth cues is one of the most important contents. Therefore, we proposed a least-square depth estimation method based on multi-scale texture features, which can better describe the relationship between the texture characteristics and the scene depth. first we captured the texture energy of different scales as the features of the image using Laws' filters to image texture gradients, texture variations and colors, respectively, and then we trained these features of the training sample sets to get the relationship parameters between texture cues and scene depth, which can finally be applied to estimate the depth of the test sample sets. The experimental results show that the method we proposed performs well in scene depth information extraction, meanwhile it requires less constraint for scene structure and space layout of 2D images. Consequently, it has good robustness because it need no specific analysis to the images with special depth cues such as linear perspective.

This paper will specifically examine scenario planning and user experience by two-stage decision-making processes by the integration of quality function deployment (QFD) and analytic hierarchy process (AHP) approach for smart TV human-computer interaction design. This research provides not only qualitative analysis but also the quantitative measurement of the priority and relevance of the features derived for smart TV HCI design.

This paper presents a real-time processing flow for ICA based EEG acquisition system with eye blink artifact elimination. Since EEG signals are one of the feeblest physiological electrical signals, it is easily contaminated by artifacts. Previously, ICA was used to extract artifacts from an EEG data segment in a time period. After processing of ICA, automatic artifact detection and elimination are performed. After that, artifact free EEG signals are reconstructed. Recently, many kinds of EEG applications such as BCIs are proposed to control machines through EEG directly. In order to make BCIs more feasible and reliable, the EEG signals used for BCIs need to be acquired from human without artifacts in real-time. In this work, a real-time ICA algorithm, ORICA, is adopted. Since eye blink artifact dose the most significant harm to EEG signals, this work focus on the automatic eye blink artifact elimination and the algorithm used for eye blink artifact detection is sample entropy. With these algorithms and the real-time processing flow we proposed, processing result of each EEG raw data is finished in 0.25 s after each sample time. In the end of this paper, the method used to evaluate the performance of eye blink artifact elimination is provided. Real EEG signals are also processed and the operation results are shown to remove the eye blink artifacts exactly without misses.

Most web-based e-learning platforms for teaching and learning can record individual action history of the access to course materials and activities. In general, the click through rate (CTR) is recorded as the action to represent the history of people who review the course online. By analyzing CTR only would exclude the important information from the visual behaviors. This study proposed a different method for collecting navigation behaviors. A method uses a web-based camera with the techniques of analyzing the source codes and tracking the objects of web pages and puts people in a more natural and undisturbed environment.

This paper presents an efficient VLSI implementation of on-line recursive ICA (ORICA) processor for real-time multi-channel EEG signal separation. The proposed design contains a system control unit, a whitening unit, a singular value decomposition unit, a floating matrix multiply unit and, and an ORICA weight training unit. Because the input sample rate of the ORICA processor is 128 Hz, the ORICA processor should produce independent components before the next sample is input in 1/128 s. Under the timing constraints of commutating multi-channel ORICA in real time, the design of the ORICA processor is a mixed architecture, which is designed as different hardware parallelism according to the complexity of processing units. The shared arithmetic processing unit and shared register can reduce hardware complexity and power consumption. The proposed design is implemented used TSMC 90 nm CMOS technology with 8-channel EEG processing in 128 Hz sample rate of raw data and consumes 2.827 mW at 50 MHz clock rate. The performance of the proposed design is also shown to reach 0.0078125 s latency after each EEG sample time, and the average correlation coefficient between the original source signals and extracted ORICA signals for each 1 s frame is 0.9763.

This paper presents a parallel VLSI architecture of a singular value decomposition (SVD) processor for real-time multi-channel electroencephalography (EEG) System. In the recent years, EEG has been widely applied on engineering research, medical diagnosis, and so on. More and more studies regarding brain-computer interface (BCI) and other related applications have been published. In order to increase the accuracy of BCI, the need for a real-time multi-channel EEG System is very urgent. Because an EEG system uses a SVD processor to calculate inverse matrix of target ones, the real-time requirement of the EEG system depends on the operation latency of the SVD processor. Moreover, the accuracy of results obtained from SVD processor directly affects the performance of the system. Generally, SVD is based on coordinate rotation digital computer (CORDIC) algorithm in hardware implementation. Therefore, there is a trade-off between the iteration number of the CORDIC engine, which is related to the computing latency of the SVD processor, and accuracy of SVD the results. In this paper, the parallel architecture of the SVD processor can efficiently shorten the clock cycle of iteration times and provide a precise inverse matrix result. This work not only upgrades the EEG system practicability, but also ensures the feasibility of real-time application. The proposed SVD processor is implemented in the 8-channel EEG system using the TSMC 90 nm CMOS technology.

This paper presents a VLSI design of singular value decomposition (SVD) processor used in real-time independent component analysis (ICA) computation for multi-channel electroencephalography (EEG) system. EEG signals are easily influenced by other artifacts. To acquire artifact free EEG signals, ICA is a popular method for artifact removal. Results obtained after the pre-processing of ICA are often used for further applications such as brain computer interfaces (BCIs). In order to improve the feasibility and convenience of BCIs, a real-time ICA pre-processing is required. Because SVD is used frequently in computations of ICA, a SVD processor used for real-time ICA computation is essential. This paper aims to develop a custom SVD for multi-channel EEG systems based on ICA. During the ICA process, the proposed processor aims to solve the inverse and inverse square root matrices in real time. And the processor obtains a highly accurate result since a novel design concept for renewing data flow and parallel data processing are provided in this research. This processor is developed with TSMC 90nm CMOS technology in an 8-channel EEG system. The performance of the proposed SVD is also provided with the processing result of the EEG system.

This paper presents an online recursive ICA (ORICA) based real-time multi-channel EEG system on chip design with automatic eye blink artifact rejection. Since EEG signals are very feeble, they are easy to be contaminated by artifacts. Among all artifacts, eye blink artifact dose the most significant harm to EEG signals. For acquiring artifact free EEG signals, ICA is a popular method for artifacts extraction. After extraction of ICA, eye blink artifacts are rejected to improve the reliability of EEG applications such as brain computer interfaces (BCIs). To promote the feasibility and convenience of BCIs, a real-time ICA algorithm, ORICA, is adopted in this system. With ORICA, the system immediately finishes each ICA result after each sample time. After each ICA result is generated, the automatic eye blink artifact rejection and inverse ICA are performed to acquire eye blink artifact free EEG signals in realtime. For system portability and high integration, a front-end interface of commercial IC, ADS1298 provided by TI, and Bluetooth interface, UART, are implemented. The system is designed used TSMC 90nm CMOS technology with 8 channels EEG processing in 128 Hz sample rate of raw data and consumes 8.56 mW at 50 MHz clock rate. The design methods of the proposed EEG system are provided in this paper. The performance and processing results of the system are also shown to reach 0.2532 s latency after each EEG sample time.

This paper presents an effective chip implementation of a real-time eight-channel electroencephalogram signal processor based on on-line recursive independent component analysis (ORICA) algorithm. The system architecture is composed of a memory unit, a whitening unit, an ORICA training unit, and an ORICA computation unit. The proposed architecture is implemented using TSMC 90 nm CMOS technology. It occupies a core area of 800×800 μm2 and consumes 4.18 mW at a core supply voltage of 1.0 V and 50 MHz clock operating frequency. Simulated super and sub-Gaussian signals are used to verify the system. The separated signals match those obtained using off-line Matlab-based analysis.

In recent years, with the popularization of higher education development and the scale of in school students is expanding, Currently in the international it is a general trend of implementing the hierarchical teaching in high school. Based on mission and the current status of the university education, with professional basis requirements of subject settings, this paper discusses practical ideas that the part of professional implements the hierarchical teaching model and disciplines building ideas of classification of teaching management. Through research summarized thus bringing the direction of change in educational administration, meanwhile, the implementation of specific policies for hierarchical teaching the corresponding hierarchical management solutions and suggestions in improving the quality of teaching to enhance students' enthusiasm and creativity, promote development of individual students has a positive reference.

This study analyzed the domestic tourism demand for household in Chinese Taipei from 2000 to 2008 with the Almost Ideal Demand System (AIDS) model. The results show that the domestic tourism demand has slack and busy seasons. Both the SARS attack in 2003 and the financial crisis in 2008 significantly affected the domestic tourism demand. Such results could be referred to by the industry in developing pricing strategies. The substitute or complementary relationship between domestic tourism demand, the tour industry can meet the tourism demand of household by offering various combinations through the substitute or complementary relationships between tour activities.

This paper uses a panel vector autoregressive model to examine the existence of political business cycles across Taiwan, Japan, and South Korea over the 1988 to 2008 period. The empirical results indicate that the electoral dummy variable has a significantly positive impact on the government expenditures. The government expenditures in the election year and the previous year of election are significantly increased. There is evidence supporting the political business cycles across Taiwan, Japan, and South Korea. Therefore, reducing public debts caused by the candidates' election promises may be a way to improve the situation of government fiscal worsening.

In this paper we use multivariate panel error correction models (PECMs) to investigate the long and short-run budgetary decision making of Taiwan and China over the 1965 to 2008 period. The empirical results show that, in the long run, for both Taiwan and China, government revenues and expenditures adjust towards budgetary balance supporting the fiscal synchronization hypothesis. In the short run, incrementalism occurs in both Taiwan and China. Although Taiwan and China have similarities in culture, language, and even in way of living, the two have vastly different government systems. Interestingly, the nature of the long run and short-run budgetary process for these two countries are surprisingly similar.

A novel laminar stream-based microfluidic chip that can serve as the high efficient technique for human sperm motility sorting through the use of flow cytometry was proposed in this study. Recently, the selection of good sperms in semen plays an important role in the reproductive medicine. However, the traditional sorting methods are normally expensive and time consuming. This modified microfluidic chip provided a new selection, which had advantages of low-cost, high-precision, and disposable. This paper was proposed and tested the interactions between laminar flow and sperm activity. Due to the laminar flow in a different direction, the semen samples were loaded into a curve laminar flow through straight-curve stream and it would offer a y component force for dead sperms to be selected. In addition, the semen samples were loaded into a straight laminar flow and the results showed that most of the proportion of motile sperms with lateral activity can swim across the two fluid boundaries to be sorted. From the results of flow cytometry analysis, the injected semen sample from a straight inlet with the counted percentage of living sperms was 92.68% by the observation in outlet reservoir. Therefore, the microfluidic chip with remarkable sperm motility sorting has the potential for reproductive medicine.

Non-equilibrium molecular dynamics simulations are performed to investigate the influences of chemisorption biphenyl rings under axial thermal loadings of 1.65 K/A. This investigation determines carbon nanotube thermal conductivity at biphenyl rings sorption density ranging from 0%~30%, using a non-equilibrium molecular dynamics simulation with true carbon potentials. The thermal impact causes system fluctuation in the initial 3 ps leading to a transport region temperature as high as 400K. The thermal relaxation process reduces impact energy influence after 30 ps and leads to Maxwell's distribution. Steady-state constant heat flux is observed after thermal equilibrium. Furthermore, the temperature curves show distinct high disturbance at initial time and linear distribution along the tube axial direction after steady-state. Simulations are performed on pristine CNTs and CNTs on which 0.25%, 1%, 5%, 10%, 15%, 20%, 25% and 30% of the carbon atoms have a bonded biphenyl group. The case where biphenyl groups are attached to 0.25% of the atoms on a nanotube is illustrated in the left panel of Figure 1. Results suggest that thermal conductivity value increases with increasing CNT subjected to thermal loading up to a temperature gradient of ~ 20 K/A representing rapidly thermal conductivity drop at sorption density of 10%. The functionalized CNTs all show significantly smaller thermal conductivities. Simulation results yield precise understanding of nano-scale transient heat transfer characteristics in a single-wall carbon nanotube.

The control of flocking motion must be effective to avoid the various obstacles, and maintain the formation to the target point. A decentralized control method is used to control intelligent groups distributed. The topology model of dynamic network of intelligent groups has been combined with artificial potential field.The control algorithm for obstacle avoidance has been improved. Limit-cycle method for flocking motion with obstacle avoidance in global unknown environment is studied. The limitations of the shape of the obstacle have been overcome. And the limit-cycle method makes the swarm avoid obstacles smoothly.