Search

China's environmental problems are not only related to the fundamental interests of the broad masses of the people, but also to China's national security and international image. At present, China's environmental protection work is facing a complex situation. Pollution sources can be divided into natural pollution sources and man-made pollution sources. Natural sources of pollution refer to places where nature releases harmful substances or causes harmful effects to the environment, such as active volcanoes. Man-made pollution source refers to the pollution source formed by human activities, and is also the main object of environmental protection research and control. Among the man-made pollution sources, air pollution sources, water pollution sources and soil pollution sources can be classified according to the main objects of pollution. Among them, air pollution sources and water pollution sources have the greatest impact on human life. Therefore, it has become an important subject worthy of in-depth discussion to take automatic and electronic measures for potential environmental pollution incidents, discover environmental pollution problems in time, reduce the probability of environmental pollution incidents, and even put some major environmental pollution incidents in their infancy. In this paper, deep learning method is used to classify the existing key pollution source video. Water pollution experiments show that the accuracy of video counting reaches 93.1%, which is better than other video processing schemes. The operation time of the system reaches acceptable range, and a solution to meet the real-time requirement is put forward.

Objective To evaluate the effectiveness and safety of rhubarb-based Chinese herbal formulae (RCHF), which are widely used to treat hepatic encephalopathy (HE) in China. Methods Nine online databases were searched from inception to November 22, 2016. Eligible studies were randomized controlled trials of Traditional Chinese Medicine (TCM) treatment for adult patients (≥ 18 years old) with HE. Outcomes such as mortality rate, clinical response rate, blood ammonia level, and alanine aminotransferase were evaluated between TCM group and control group. Results Thirty studies involving 2661 HE patients were analyzed. Most studies used RCHF treatment. Compared with conventional treatment as usual, lactulose, and vinegar, RCHF were associated with significant improvement in clinical response rate [risk ratio (RR) = 1.33, 95% confidence interval (CI) = 1.25, 1.43, I2 = 0%; RR = 1.26, 95% CI = 1.14, 1.38, I2 = 22%; and RR = 1.19, 95% CI = 1.06, 1.33, I2 = 0%, respectively] and significant reductions in levels of blood ammonia and alanine aminotransferase. Only minor RCHF-associated adverse events, such as abdominal pain (0.3%), anal tenesmus (0.3%), and diarrhea (2.3%), were reported, and there were no significant differences in these events between the treatment group and the three types of control group. Conclusion The findings suggest that RCHF may be an alternative treatment option for HE patients. More rigorous multicenter studies with larger samples and longer observational periods are needed to confirm these findings.

Objective To compare the causes, clinical symptoms, laboratory test results, and prognosis in patients with acute liver failure (ALF) induced by traditional Chinese medicines (TCM) and by Western Medicines (WM). Methods The medical histories of patients who were diagnosed with drug-induced ALF (DALF) (n = 96) after hospitalization in the 302 Military Hospital between January 2010 and December 2015 were retrospectively examined. Results Fifty-eight of the 96 DALF patients (60.4%) had a hepatocellular pattern of DALF, 16 patients (16.7%) had a cholestatic pattern, and 22 patients (22.9%) had a mixed pattern. DALF resolved in 24 patients (25.0%). Twenty-five patients (26.0%) developed chronic liver injury, 43 patients (44.8%) died, and 4 patients (4.2%) underwent liver transplantation. There were 42 ALF patients (43.8%) who received WM, and 32 ALF patients (33.3%) who received TCM. TCM-induced ALF patients had a higher average age [42.4 ± 18.4) vs (33.5 ± 17.9) years, P = 0.04] and higher creatinine and urine nitrogen levels [(155.2 ± 108.8) vs (97.5 ± 130.4) mmol/L, P = 0.047; (9.1 ± 7.7) vs (4.3 ± 5.0) mmol/L, P = 0.002, respectively]. Patients with TCM-induced ALF exhibited an increased risk of renal injury [odds ratio (OR), 3.75; 95% confidence interval (CI), 1.330-10.577]. The 14 patients with TCM-induced ALF who died exhibited higher creatinine levels than the 18 patients with TCM-induced ALF patients who survived [(218.7 ± 111.6) vs (105.8 ± 78.4) mmol/L, P = 0.002]. They were also more likely to exhibit ascites (85.7% vs 44.4%, P = 0.017) and hepatorenal syndrome (78.6% vs 22.2%, P = 0.002). Conclusion TCM-induced ALF was more likely to be accompanied by renal injury than was WM-induced ALF, especially in TCM-induced ALF patients who died.

As the technological progress of mobile Internet, smartphone based on Android OS accounts for the vast majority of market share. The traditional encryption technology cannot resolve the dilemma in smartphone information leakage, and the Android-based authentication system in view of biometric recognition emerge to offer more reliable information assurance. In this paper, we summarize several biometrics providing their attributes. Furthermore, we also review the algorithmic framework and performance index acting on authentication techniques. Thus, typical identity authentication systems including their experimental results are concluded and analyzed in the survey. The article is written with an intention to provide an in-depth overview of Android-based biometric verification systems to the readers.

In the modern era of rapid development of information technology and mobile Internet, mobile terminals represented by Android smart phones are applied to various fields of life, and the number of users is far more than that of traditional PC terminals. Mobile financial applications marked by mobile banking and mobile payment have higher access control requirements, while traditional text passwords, dynamic passwords and verification codes have hidden dangers of being stolen and copied. How to apply biometrics to mobile identity authentication with high security requirements has become a hot issue for scholars all over the world. Lip reading and speech biometrics are studied in this paper, and a multi-biometric fusion authentication system based on both on Android smart phone platform is developed.

UAV remote sensing has been widely used in emergency rescue, disaster relief, environmental monitoring, urban planning, and so on. Image recognition and image location in environmental monitoring has become an academic hotspot in the field of computer vision. Convolution neural network model is the most commonly used image processing model. Compared with the traditional artificial neural network model, convolution neural network has more hidden layers. Its unique convolution and pooling operations have higher efficiency in image processing. It has incomparable advantages in image recognition and location and other forms of two-dimensional graphics tasks. As a new deformation of convolution neural network, residual neural network aims to make convolution layer learn a kind of residual instead of a direct learning goal. After analyzing the characteristics of CNN model for image feature representation and residual network, a residual network model is built. The UAV remote sensing system is selected as the platform to acquire image data, and the problem of image recognition based on residual neural network is studied, which is verified by experiment simulation and precision analysis. Finally, the problems and experiences in the process of learning and designing are discussed, and the future improvements in the field of image target location and recognition are prospected.

With the wide use of lithium battery, its online monitoring and residual capacity prediction have been paid much attention. There are two methods for predicting the residual capacity of lithium batteries, namely model method and data-driven method. The traditional model method requires in-depth understanding of the material characteristics and aging mechanism of the battery. However, it is difficult to establish an accurate model due to the complex electrochemical reactions in the battery and the vulnerability to external factors. The data-driven law has been applied more widely because of its good applicability and flexibility. This paper presents a method of battery capacity prediction based on fuzzy logic and neural networks. The lithium battery data published by PCoE are selected for the test, and the results show that the prediction error of the method for the residual capacity of single battery is less than 2%, which indicates that the method has a good applicability for the complex nonlinear system of lithium battery pack, and can obtain accurate battery capacity prediction value, and it has a good application prospect.

China is one of the countries with the most natural disasters in the world. Natural disasters seriously endanger people’s safety and economic situations. Therefore, it is very necessary to strengthen the prediction and research on the characteristics and the causes of various natural disasters. Based on the Internet distributed and incremental crawling technology, this system crawls the knowledge literatures, disaster news and professional disaster data from domestic and overseas on natural disasters, as well as drawing on the advanced knowledge management concept, the captured data will be displayed in the form of knowledge tree. Integrating the authoritative global typhoon live maps and orthophotos which acquired by UAV to better monitor and predict typhoons; This system determined that the SSM framework is used as the system framework, combined MySQL and Redis database to store data stably and efficiently. In addition, the Echarts component can be used to analyze and visualize the crawled data. Big data analysis techniques such as knowledge graphs and related algorithms for literature analysis, data cleaning to better data store and display. In the end, it will enable the researchers to more easily and efficiently carry out the knowledge sharing, collaborative cooperation and interactive communication, which has the certain research significance and use value.

With the steady growth of the economy and the rapid development of modern industrial technology, the problem of environmental pollution has increased. To continue to develop, it is necessary to thoroughly implement the sustainable development strategy, and we must pay more attention to environmental issues. One of the important management tools implemented in China for environmental management is environmental quality monitoring and evaluation. Environmental quality monitoring can scientifically evaluate the environmental quality of a region, scientifically evaluate and forecast the environmental management and environmental engineering, and provide scientific basis for environmental management, environmental engineering, formulation of environmental standards, environmental planning, comprehensive prevention and control of environmental pollution, and ecological environment construction. This paper will discuss the basic principles of neural network and the implementation process of MATLAB and in the MATLAB software implementation and display process. At the same time, the results of different parameters are analyzed through experiments, and the network parameters are constantly adjusted to improve the accuracy of the evaluation results. Taking the regional environment as an example, two monitoring methods are proposed, and a variety of neural network models are used to analyze each prediction method. Case study results show that the latter method has a better prediction effect.

Recently, increasing attentions have been concentrated upon the judgement of video-based traffic accidents, which serves as a crucial technology of intelligent transportation system (ITS). This paper introduces a novel vehicle tracking method based on projection extreme learning machine (PELM) by taking consideration of poor performance in the dealing with high-dimensional and low-samples. Firstly, samples near the target in the first frame of a video is employed as training samples to the PELM classifier. The algorithm uses projection vectors instead of random assignment as the weights of the input layer. Then the PELM classifier is trained with the samples near the target in the following frames to update parameters and obtain the target location with the maximum response. Finally, the fitting error and maximum-range trajectory are adopted to judge whether the vehicle is collided or not. Simulation experiments have shown the proposed method can achieve quite satisfying results.

Motivated by the security problems caused by the shortage of reinforcement quantity and the sparse distribution of steel bars, this paper proposes a novel steel bars counting method based on machine vision which focus on processing videos and images to obtain the quantity of steel bars. Firstly, we use image capture technology to convert the video into an image. Then preprocessing and edge detection of the above images are used to set up the region of interest through template matching. Finally, we use the Hough transform line detection technique to extract the lines and get the number of bars in the calibration area or in the entire video. Experiments show that the accuracy ratio of steel bars counting is up to 100%, and at the same time this method can detect a many steel bars in the video rapidly, reducing the workload of people.

The research on classification method of malicious code is helpful for researchers to understand attack characteristics quickly, and help to reduce the loss of users and even the states. Currently, most of the malware classification methods are based on supervised learning algorithms, but it is powerless for the small number of labeled samples. Therefore, in this paper, we propose a new malware classification method, which is based on semi-supervised learning algorithm. First, we extract the impactful static features and dynamic features to serialize and obtain features of high dimension. Then, we select them with Ensemble Feature Grader consistent with Information Gain, Random Forest and Logistic Regression with \(L_1\) and \(L_2\), and reduce dimension again with PCA. Finally, we use Learning with local and global consistency algorithm with K-means to classify malwares. The experimental results of comparison among SVM, LLGC and K-means + LLGC show that using of the feature extraction, feature reduction and classification method, K-means + LLGC algorithm is superior to LLGC in both classification accuracy and efficiency, the accuracy is increased by 2% to 3%, and the accuracy is more than SVM when the number of labeled samples is small.

Radio Frequency Identification (RFID) is a technology to exchange messages by radio frequency (RF) between Reader and Tag. As a RFID test system, stability, consistency and adaptation are very important. At a certain budget, the above properties of the test system will be limited. In order to guarantee the performance of the system and get test results as accurate as possible, a lot of solutions should be applied to the whole system. Calibration is a very important part in every system. By proper calibration method, we can get a system with high precision, good stability and consistency. Furthermore, with calibration by hardware of high precision, all the test uncertainties will be able to trace back to that system. Certainly, the system used for calibration will be calibrated by another better system. In such way, we can cost down the test system and at the same time make sure the correctness of the test results. There are two ways of calibration, manually and automatically. Manual calibration requires mass labor which will absolutely raise the cost of total system. If the calibration could be done the same time the RFID system is testing, this will be the fast and most accurate way. In this paper, how to guarantee the accuracy of RFID system will be discussed, especially calibration.

As the number of vehicles increases significantly, traffic congestion has become a major social problem in recent years. Such a situation can be alleviated effectively with the emerging of vehicular ad hoc network (VANET)-based real-time path planning systems. However, existing systems face the challenges of poor anti-congestion capability and high complexity. To address the related issues, an road side unit (RSU)-based architecture is proposed in this paper, in which a city is divided into different areas, each with an RSU. Based on the architecture, a decentralized and hierarchical real-time path planning algorithm is proposed. The path planning problem is formulated from two layers, i.e., area path selection in upper layer and intra-area routing in bottom layer, both of which target to minimize the average travel time. Numerical results show that, our proposed algorithm inherits the anti-congestion capability and owns the advantage of low complexity, as compared with the shortest path algorithm and centralized algorithm.

This paper presents a high-responsivity and high-speed InGaAs/InP PIN photodetector integrated onto the silicon waveguide substrate utilizing the divinyltetramethyldisiloxane-benzocyclobutene (DVS-BCB) adhesive bonding method. A grating coupler is adopted to couple light from the fiber to the silicon waveguide. Light in the silicon photonic waveguide is evanescently coupled into the photodetector. The integrated photodetector structure is first simulated using the FDTD (finite difference time domain) solutions software and the simulation results show a detection efficiency of 95%. According to the simulation result, the integrated photodetector is fabricated. The measured responsivity of the fabricated integrated photodetector with a detection length of 30 μm is 0.89 A/W excluding the coupling loss between the fiber and the grating coupler and the silicon propagation loss at the wavelength of 1550 nm with a reverse bias voltage of 3 V. Measured 3-dB bandwidth is 27 GHz using the Lightwave Component Analyzer (LCA). The eye diagram signal test results indicate that the photodetector can operate at a high speed of 40 Gbit/s. The integrated photodetector is of great significance in the silicon-based optoelectronic integrated chip which can be applied to the optical communication and the super node data transmission chip of the high-performance computer.

In LLL images, there is a lot of random flicker noise. When denoising in LLL image, the traditional time-domain recursive noise reduction will bring smear in moving images. An improved algorithm is needed. To improve the time-domain recursive filter, the block that contain moving target in images should not participate the recursive algorithm. Time-spatial recursive denoising for LLL images based on motion detection is proposed to improve the time-domain recursive noise reduction. The blocks that contain moving object and background are processed respectively. Finally, simulation results show the effectiveness of the proposed method.

In this paper, we present a football event detection method by using multiple feature extraction and fusion. Instead of using low-level features, the proposed method is built upon visual, auditory features, text and audio keywords. Promising event detection results have been achieved. By using the proposed method, we have been able to detect the football events accurately. Experimental results have shown our proposed method is effective for football event detection.

This paper proposes a predictive controller based on adaptive resource management techniques for multi-service wireless cellular networks to keep the handoff failure rate (HFR) below a target value regardless of the traffic dynamics and bandwidth requirements of the service types. In this paper, we propose a neural nod feedback algorithm to control the reserved resource. Simulations results indicate that this algorithm can maintain target HFR for any service type even when the traffic load rapidly fluctuates and system parameters vary over a wide range. This algorithm has superior performance in terms of total carried traffic.

Semantic sports video analysis has attracted more and more attention recently. In this paper, we present a football event detection method by using multiple feature extraction and fusion. Instead of using low-level features, the proposed method is built upon visual, auditory features, text and audio keywords. Promising event detection results have been achieved. By using the proposed method, we have been able to detect the football events accurately. Experimental results have shown our proposed method is effective for football event detection.

Third generation city-wide distributed surveillance systems are built upon networks of sensors, actuators and computation nodes. In such large systems, different groups of nodes may belong to separate parties who want to control access to the resources provided by their nodes. The access control mechanism must verify the identities of nodes and check their authorization for accessing resources. The system must also support dynamic group reconfiguration mandated by factors such as changes in organizational structure or confidentiality rules and recover gracefully from security breaches. This paper presents a mechanism based on Public Key Infrastructure (PKI) and certificate hierarchies for fulfilling these requirements in a secure communication platform supporting both synchronous and asynchronous operations in a distributed surveillance system. Dynamic reconfiguration of groups and expulsion of compromised parts of the system is implemented by utilizing certificate revocation.