Abstract:Based on the research progress of Simultaneous Localization And Mapping (SLAM), this paper systematically summarizes the research on SLAM of mobile robot by reviewing the research methods of SLAM in the last thirty years and points out the three key problems. In view of these three problems, the methods of SLAM based on probability estimation and vision are introduced. Compared and summarized the SLAM method based on probability estimation, the development of vision-based SLAM method based on different characteristics of vision sensors is elaborated, and the development of vision-based SLAM method is described in detail. By analyzing the characteristics of different kinds of vision sensors, the pros and cons of each method are compared and analyzed, and the existing problems are discussed. Finally, the future development directions of SLAM are presented.

Abstract:SIFT feature has been widely used in the homologous copy-paste forgery detection. Due to the rejection of color information, it results in some mismatching of key points, so we propose another method based on Colored Scale Invariant Feature Transform (CSIFT), a feature combining variable color information with Scale Invariant Feature Transform (SIFT) when extracting feature points decreasing the possibility of false matching and greatly reducing the time of feature point extraction and matching. In this study, we firstly segment video by Structural Similarity Index Measurement (SSIM) and extract the first frame of every part as the key frame. Then, we extract CSIFT feature points of the key frame and match feature points. After that, we locate copy-paste areas. The final step is a target tracking algorithm used to calculate the copy-paste areas in the following frame sequence. The robustness and efficiency of the algorithm is verified by experiments. Compared with algorithms based on SIFT, the proposed method has higher time efficiency and accuracy.

Abstract:In the operation room, the traditional human-computer interaction technology relies heavily on the touch-based devices such as mouse, keyboard, and touch screen, which raises the risk of infection during the surgery. Human-computer interaction based on gesture recognition has a great advantage in the cost and interactive naturalness. This paper introduces a gesture-based control technology utilizing Leap Motion, the use of AR technology and Leap Motion equipment make doctors do not need to touch the surgical equipment during surgery. With such help of non-contact equipment, doctors can be completely away from the interference and surgical risk because of the contact to the equipment during surgery. The Leap Motion based interaction technology provides the possibility of reducing the risk of surgical infection in the operation room.

Abstract:In recent years, the number of malware on the Android platform has a geometrical growth. Therefore, it is very necessary to have a method to detect Android malware. This study experiments with a large number of Android malware samples and machine learning technology to establish a prediction model for malware classification, which is run in the static detection process. First, we obtain the permissions and the dangerous API information of Android applications, the permissions feature in its AndroidManifest.xml file by decompiling APK files and its dangerous API features by translating decompiles class.dex files into smali files together with the baksmali tool. Then, we use multiple classification algorithms and preprocessing algorithm to compare the accuracy rate of the single detection and the conjoint detection. The experimental results show that the accuracy rate of the conjoint detection is higher than that of the single detection, and the accuracy rate reaches up to 97.5%.

Abstract:Hesitance Fuzzy Linguistic Term Sets (HFLTSs) allow decision makers to evaluate a property in several possible linguistic terms. Recently, HFLTSs based fuzzy clustering analysis draws increasing attention. Considering that the current fuzzy clustering algorithm based on HFLTSs still costs large computation, this study proposes a novel orthogonal fuzzy clustering algorithm. Firstly, calculate the distance measures between samples to construct distance measure matrix, and then calculate the matrix's equivalent matrix. Secondly, cut the equivalent matrix according to its confidence level to obtain the corresponding cutting matrix. Finally, obtain the clustering result based on the orthogonal relationship between the column vectors of the cutting matrix. This algorithm has simple steps and low computational complexity. It is also suitable for large-scale fuzzy clustering problems. At last, the feasibility and efficiency of this algorithm are proved by a practical application with k-means clustering algorithm.

Abstract:In recent years, the rapid development of deep learning has led more and more people to engage in related research work. However, many researchers construct deep neural network models based on standard algorithms or improved algorithms, but do not understand the algorithm itself and the factors that affect the performance of the model, resulting in more or less blind application in many applications. By studying the deep neural network, the activation function of the important influencing factors was studied. First, the activation function is analyzed to influence the depth neural network. Then, the development of activation function and the principle and performance of different activation functions are analyzed and summarized. Finally, based on the Caffe framework, the CNN is used to classify and identify MNIST data sets. Five kinds of commonly used activation functions are analyzed and compared comprehensively to provide a reference for the selection of activation function in the design of deep neural network model.

Abstract:A Cyber-Physical System (CPS) is a mixed system integrated with computation, communication, and physical processes, which takes an important position in the smart city, and there are many challenges in its security issues. Firstly, we establish a cyber security model for CPS, providing with the security threats of various components of CPS. Then, we propose a CPS threats analysis method and provide experimental test results. The results show that this method can achieve rapid large-scale security threat modeling and automatic analysis. Consequently, it can provide technical supports for the security assurance of the key information infrastructures in smart cities. Finally, we summarized the research advance and future research direction of security threats on CPS in the smart city.

Abstract:Due to any network failure of the power dispatching network may lead to serious accidents, it is required to have high reliability and safety. Faced the large amount of data, the traditional network monitoring system cannot meet the demand at present in terms of the actual processing capacity and expansion capacity. Therefore, a special real-time data analysis platform is needed for the real-time analysis and processing of large amount of data. This study constructed a platform based on flow calculation. Spark Streaming in Apache Spark of the open source stream computing framework, adding Kafka message queue and Redis memory database components, provides stable data sources and efficient interface for data analysis and data service platform, so as to realize the real-time analysis and processing of all kinds of massive data, thus to complete the flow anomaly monitoring of power dispatching network.

Abstract:The improper sitting posture of reading and writing can have a serious effect on the health and learning efficiency of the young. This study designed a new type of portable, real-time monitoring, and intelligent reminder system of the writing and the reading sitting posture. According to the biomechanics and ergonomics rules, the system uses 40 cm×40 cm flexible force-sensitive sensors to collect sitting pressure distribution data, combined with environmental parameters to analyze results in real time. Meanwhile, the sitting position and environment change are displayed during the monitoring period in real time via the mobile phone APP. The system is easy to understand and interactive, which can help teenagers to correct improper sitting postures. With the convenience of mobile phone, it is convenient for users to achieve real-time warning and analysis.

Abstract:Email is today's important communication tool, but it is also one of the main ways of cyber attack. As a result of certificates mistakenly issued by CA agency, man-in-the-middle downgrade attack, and the proposal of DNS-based Authentication of Named Entities (DANE), new progress has been made on the improvement of the current email protocol and the security of email. This study combs the widely used email protocol from the point of view of email encryption and verification, analyzes its advantages and disadvantages, summarizes the latest research progress of email protocols and the improvement of current email protocol, and proposes a secure email system architecture based on DANE. Finally, the development direction of DANE-based email system is summarized and prospected.

Abstract:EAST device is the world's first non-circular cross-section superconducting Tokamak nuclear fusion experimental device, and has developed into an important international cooperation experimental platform. In order to expand and facilitate the cooperation unit to participate in the experiment, the development of EAST remote plasma control system is proposed, the system uses Web development model, functioning to provide remote customers with access to experimental data, real-time control services, and discharge program settings services. The portal system design is mainly responsible for authentication and authorization, request format inspection and technical data check, and other security features. Authentication and authorization phase use three-stage verification, the main technology includes VPN, digital certificates, random verification code, and other network security technology. Request format check and technical data check use modular technology, for different discharge program, a separate module is written to ensure system scalability.

Abstract:Enterprise networks that were built with STP and VRRP are difficult to deploy, difficult to expand, and are of poor availability and reliability. This article analyzes the root cause of these problems and presents a solution to construct enterprise network by using virtualization technology of cloud computing network equipment, i.e., build a stable and reliable enterprise network through effective integration of the simulation experiment, the virtualization technology, and dynamic link aggregation and virtual splitting detection mechanism. In order to evaluate the performance of the scheme, the article gives a number of comparative analysis of the availability, manageability, scalability, and other aspect, and proves that the enterprise network built by virtualization technology based on cloud computing is superior to the network by traditional way in terms of stability, robustness, and other aspects.

Abstract:The Internet has a huge amount of data, someone often need to acquire structural data of multiple source station to support data analysis, disinterment. The artificial cost of different customized website data acquisition program is very high. This paper presented a scheme of automatic data structuring in web crawler. Taking an e-commerce website as an example, this paper confirmed the feasibility of structured extraction scheme from the theoretical point of view based on its unified hierarchical structure, vertical domain, and data corpus. This study proposed the similar duplicate detection and attribute based semantic label matching algorithm, implemented analyzing the structure and matching the target fields, and designed a preset matching template and the reuse mechanism of structural analysis results, for management and tuning the system. Practical application and error rate test show that this scheme is very feasible and can greatly reduce artificial coding, and the error rate is low. The design idea can be applied to the subject crawler system in other fields, and quickly obtain large amount of data from many sites, and let people focus more on structured data processing and information disinterment.

Abstract:Android Software Development Kit (SDK), a collection of Android tools, is an effective Java-based kit to develop Android APPlication (APP). However, Android APP's performance is usually limited by the backend Dalvik virtual machine environment. A novel practical model that Android APP developed by C/C++ codes is proposed, because Java Native Interface (JNI) can support C/C++ dynamic library callback mechanism. This paper firstly analyzes and compares the characters between the general development mode for Android APP and Qt cross-platform development framework, and presents a novel principle and a technical relationship between Android APP and Qt framework. Then, it describes the main processes, abstract implementation steps, and software list for the solution. Finally, the development of a sample APP, Mechanical CAD Teaching Assistant, is given as an example to verify the effectiveness and feasibility of the development pattern. This work provides a new way to use C/C++ programming language to develop APPs running on Android operating system.

Abstract:With the rapid development of economy and society, environmental pollution especially haze and toxic gases pollution do great harm to people's health and are getting more and more serious. Thus the demands for environment information, such as PM2.5, temperature, and humidity, and all kinds of gas concentration monitoring are becoming higher and higher. Therefore, it is of great significance to develop and design an environment monitoring data collection system with various functions, high reliability, and good portability. The study uses sensor and embedded technologies to design a new kind of portable comprehensive environmental monitoring data acquisition system. The experimental results and error analysis show that the data acquisition system can realize real-time data collection and display function, and ensure the credibility of system data, indoor environment monitoring, coal gas, natural gas, and toxic gas monitoring.

Abstract:The purpose of this study is to solve the problems existing in the traditional parking lots, such as inefficient parking, inconvenient parking, and so on, using IoT-related technologies, combining software and hardware to realize the parking automation management system. In the parking lot, ultrasonic parking sensors check parking status and transmit parking status information to the host computer through ZigBee wireless network. According to the actual parking status information, the upper computer uses ant colony algorithm to calculate the best parking route for best parking space. Deadlock, stagnant or even no solution situation is easy to appear in the traditional ant colony algorithm's solving process. Therefore, in this study, the traditional ant colony algorithm is improved according to the structural model established by the parking lot in solving the optimal parking space in parking lot. Finally, according to best parking route, the upper computer via TCP/IP network protocol sends the instructions of left, front, or right to display to the driver to the best parking position.

Abstract:A fatigue driving detection algorithm based on deep learning was proposed to solve the problem of poor practicability or low accuracy of existing fatigue driving detection algorithms. First, the Histogram of Oriented Gradient (HOG) feature operator was used to detect the presence of human faces. Secondly, the feature points model was used to realize the face alignment and segment the eye and mouth regions. Finally, the fatigue features of the driver's eyes were extracted by the deep convolutional neural network, and the fatigue features of the driver's mouth were fused to carry out fatigue warning. Experiments show that the proposed method can achieve significant performance improvement in terms of accuracy and real-time performance of fatigue driving detection.

Abstract:Real-time accurate prediction of short-term railway passenger demand can provide basis for real-time adjustment of passenger service structure. Railway passenger flow data has characteristics such as time-varying, nonlinear and stochastic volatility. Traditional forecasting models cannot predict the short-term passenger traffic volume accurately. This study proposed a hybrid deep learning model based on Wavelet Packet Analysis and Long Short-Term Memory (WPA-LSTM). Firstly, the original passenger volume time series is decomposed into several low-frequency and high-frequency sequences with different scales by wavelet packet. Then, the LSTM model training and prediction are carried out respectively for each sub-sequence. Finally, the prediction values of each sub-sequence are superimposed as the output of WPD-LSTM model. The model was validated by the daily passenger flow data of 367 days of one high-speed railway. The model was compared with the seasonal model and the empirical model. The experimental results show that WPD-LSTM model can effectively improve the accuracy of railway passenger traffic forecasting.

Abstract:The real-time interception of user chat content in live broadcast system is of great significance. In order to improve the accuracy and efficiency of the classification, a text classification model based on the combination of Doc2Vec and SVM is proposed to classify the chat content and judge whether the chat content should be intercepted. The First part uses the Doc2Vec model to represent the chat content as a dense numeric vector, and then an SVM classifier is used to classify. The experimental results show that the model greatly reduces the dimension of text representation with high efficiency, and it has excellent accuracy rate (97%) and recall rate (89.82%), which are superior to Naive Bayes and the logistic based on Doc2Vec.

Abstract:Natural language processing techniques have been confirmed to well classify the alarm messages based on the user intents. In this study, we propose an intelligent classification algorithm for burst alarm events. First, we analyze the features of user feedback text data. We then present a semi-automatic semantic mapping tool to calculate the probability, so as to form a valid alarm category. By application this method into the enterprise fault location, it takes advantages of (1) releasing alarms upon different user intents; (2) reducing the alarm miscalculation; and (3) helping game operation personnel accurately and quickly locating the reasons of the fault. The experimental results show that the intelligent classification algorithm based on semantic mapping can effectively improve the performance of the alarm distribution, and achieves better results than the simply algorithm for text categorization.

Abstract:As a decision tree generated algorithm, C4.5 algorithm is very influential. But the decision tree classification by C4.5 algorithm is of less accuracy, more branches, and larger scale. To solve these problems, we propose a C4.5 improved algorithm based on rough set theory and CAIM criterion. The algorithm uses the discretization method based on CAIM criterion to process the continuous attributes, which decreases the information loss degree and improve the classification accuracy in discretization. The discretized sample is reduced by attribute reduction method based on rough set theory, which eliminates the redundant attribute and trims the size of decision tree. Experiments show that the algorithm can effectively improve the classification accuracy of decision tree generated by C4.5 algorithm and reduce the scale of decision tree.

Abstract:Research on one kind of multiple traveling salesmen problem requires minimizing the maximum length of the cycles traveled by all the salesmen. A new local search operator is designed, this operator not only optimizes one cycle, but also reorganizes and optimizes two cycles, its optimization ability is better than the existing local search operators under the equivalent calculation cost. Based on this operator, an iterative strategy named "search-select-mutate- search" is proposed, according to this strategy, the Competitive Search Algorithm (CSA) is designed. The experiments on the open data sets showed that CSA performs better than recent literatures.

Abstract:In order to reduce the complexity of intra encoding in High Efficiency Video Coding (HEVC), a fast algorithm for intra encoding based on coded block texture features is proposed. Firstly, the texture complexity and direction characteristics are extracted from the current Coding Unit (CU) by using a preprocessing method. Secondly, according to the texture complexity, the algorithm can determine whether to divide some CU, skip the complex rate of distortion calculation, and cut down the CU unnecessary division and cutting. Then, according to the texture direction, the algorithm can decrease the number of intra prediction modes, which can reduce the computational complexity of intra prediction. On the basis of the experimental results, compared with HM10.0, the proposed algorithm can save 56.18% coding time on average, with the Bit Rate (BR) increases of 0.649%, and the Peak Signal-to-Noise Ratio (PSNR) losses lower 0.059 dB, respectively.

Abstract:Flavonoids are the secondary products of liquid fermentation of Phellinus igniarius and have important medical value. In this study, a hybrid intelligent algorithm combining Particle Swarm Optimization (PSO) and BP neural network is proposed to optimize the experimental environment of fermentation of Phellinus igniarius and to improve the flavonoids yield. The BP neural network is trained based on the 25 groups of experimental data and as the prediction model of flavonoid production. The experiment is compared with the mathematical regression model in the traditional response surface methodology to predict the accuracy increased by 15%. The BP neural network prediction model was used as an evaluation function in combination with PSO algorithm to optimize the experimental environment. According to the data simulation experiment, the best culture conditions of the liquid fermentation of Phellinus igniarius were obtained. The yield of Phellinus flavonoids from the previous about 1532.83 μg/mL to about 1896.4 μg/mL, yield increased by about 23.72%.

Abstract:Keyword extraction technology is the foundation of corpus construction, text analysis, and information retrieval. The traditional TFIDF algorithm is mainly based on word frequency weighting to extract keywords without considering the influence of text features. The excessive reliance on word frequency leads to the inaccuracy of extract keywords. To solve this problem, an improved algorithm has been proposed, which use the word position and the word information as factors to recalculate the weight, then we implement it in Python. Experiment shows that using this method to extract keywords can improve the recall rate, accuracy, and F-measure.

Abstract:Aiming at the high computational complexity of recursive quad-tree partition of Coding Unit (CU) in High Efficiency Video Coding (HEVC), an improved fast CU partition algorithm of HEVC was proposed. Firstly, based on the temporal-spatial correlation among neighboring CUs, the best partition structure from the co-located position of the previous frames is extracted to predict the depth of partition CU. Secondly, a fast CU partitioning structure decision algorithms are improved to reduce traversal of Coding Tree Unit (CTU) quad-tree structure. It means that the traverse operation begins from the half depths of CU partitions structure, and whether to early termination of traversal is judged before each step traversal. We compare to baseline works in HM 15.0, showing the proposed algorithm can reduce the encoding time by 55.4% while improving the coding performance and improve the coding efficiency of HEVC compared with the quad-tree recursive traversal algorithm.

Abstract:The Ant Colony Algorithm (ACO) can be applied to cluster scheduling better, but its traditional pheromone update method brings the performance matching and load balancing and so on, which affects the performance of cluster scheduling. In order to solve these problems, an Improved ACO (IACO) is put forward. The pheromone is adjusted more reasonably by adjusting the performance matching factor and the load balancing factor. It is observed that the processing time is shortened, the CPU utilization is improved, and the performance of the cluster is improved effectively.

Abstract:Aiming at solving the problem that some image encryption algorithms with the permutation-diffusion structure have low degree of encryption, this paper presents a new image encryption method. In the scrambling stage, firstly global scrambling uses gray code transformation, then scrambles pixel among row and column by chaos sequence. Diffusion part adopts the forward and reverse exclusive OR operation. The experimental results show that the proposed algorithm enhances the security of the cipher-image, key sensitivity, and increases the size of the key space, at the same time it can resist the statistical analysis, brute force attack.

Abstract:This paper presents a new cascade architecture to calculate the similarity between concepts in the ontology. In the first stage of proposed model, we use path-based methods to calculate the concept of path score in the ontology. In the second stage, we use Information Content (IC)-based methods to obtain similarity scores of two extended concepts with further consideration of the two concepts of public parent set and public subset. In the third stage, we adopt a feature integration strategy to combine all the similarity scores derived from the ontology to construct various kinds of features to characterize each concept pair and using weights to balance the concept of first-stage-scores with the third-stage-scores. In the end, BP neural network is used to obtain the similarity of two concepts. This model has been evaluated and compared with existing methods when applied to the task of semantic similarity estimation. Experimental results show that the proposed method effectively improves the accuracy and scientificity of similar calculation.

Abstract:Network security has become a priority research field in China in recent years, for there would be no national security without network security. In consideration of the characteristics of data source in network security like diversity and complexity, this study proposes an evaluation model of network security situation based on Support Vector Machine (SVM) and self-adaptive weight. The model is composed of training module and prediction module. The training module is used to obtain the key data concerned by network security through prior knowledge method and build evaluation model with a combination of SVM and weight strategy. The prediction module is used for real-time network security situation evaluation. Analysis of experimental process and results indicates that the proposed model can favorably support the real-time prediction and evaluation of the security situation of small-size networks.

Abstract:A preprocessing algorithm for stator temperature based on wind power SCADA data has been put forward in view of deviant status of wind turbine, such as insufficiency on analytical process of performance of wind turbine, the predicting inaccuracy, and the deficiency of economic benefit. The maintaining efficiency of generator stator has been improved since the analysis on data gathered from each part by SCADA system of wind power. For the temperature of stator within the generator, the process and analysis of data have been optimized, with the amelioration of Optimal Interclass Variance (OIV) algorithm and the successful monitoring the trends of temperature of stator and its abnormal temperature status. The improved optimal interclass variance algorithm has been proved feasible and efficient, which is capable of processing date of generator stator's temperature curve data and making predictions via neural networks, while improving predicting accuracy of generator stator temperature significantly.

Abstract:With the continuous development of information technology, the forecast of the recent development of things based on the network data has become a hotspot. In order to predict the monthly number of tourists in Beijing, this study established three kinds of single models with the search index of the relevant network keywords as independent variables:BP neural network, support vector regression, and random forest, and constructed a variety of combinatorial models to improve the prediction accuracy. The experimental results show that the combination of models based on GBDT have achieved higher prediction accuracy, the error is 3.16%. The forecast results can provide decision support for tourism management.

Abstract:An adaptive learning rate improvement method, based on weight change, is proposed to improve the learning rate of traditional neural network in this study. If the learning rate is too large or too small, neural network is too difficult or too slow to converge. To offset this disadvantage, the study put forward a new learning rate, based on weight gradient, to improve the convergence rate and improve the traditional neural network learning rate affected by the human experienced factors, and combined with normal distribution and gradient rise method, to size up error accuracy and convergence speed. Taking BP neural network as an example, comparing the fixed learning rate neural network, and applying classical XOR problem simulation, we verify the proposed method. The results show that this improved neural network has faster convergence speed and smaller error.

Abstract:The automatic updating method of Linux device drivers is studied and the corresponding assisted updating model based on source codes is built while a set of tools to support automatic updating of drivers are designed and implemented. This set of tools includes a tool to extract kernel-dependent interfaces for the Linux device drivers, a tool to analyze differences between kernel-dependent interfaces for two versions of kernels, and a tool to generate helpful information for updating of device drivers. Related prototypes have been tested and the results show that it can effectively improve the development and maintenance of device drivers. In addition, the concepts and calculation methods for the rate of false reporting and the rate of missing reporting about helpful information for updating are put forward, which are to be used for the quantitative indicators to evaluate assisted updating for device drivers.

Abstract:SVM classification result on imbalance data set is partial to majority class. It makes the F1_measure value of minority class inadequate. This paper presents a method which improves the classification accuracy of minority class. The method generates weights to optimize parameters b of the classification hyper plane without changing the number of samples, and combines the total number of samples, the number of support vector, the accuracy of minority class and majority class. Finally, the effectiveness of the method is proved by experiments.

Abstract:Aiming at the problem of high power consumption and low overall performance, a new design method of high speed data exchange node for Internet of Things (IoT) is proposed. The topology of IoT at the node is designed, and the modulation module, codeword addition module, and demodulation module are designed. Make each node IP and high speed data exchange module through the input port and the output port and the switching node, set a buffer queue at each input port, read data through the modulation module, transmit to code adder module for the addition operation. Results will be sent to each demodulation module after processing, and the data is transmitted to the destination IP module. The software design of high speed data exchange node is realized by Code Division Multiple Access (CDMA) technology. The experimental results show that the proposed method has high bandwidth utilization, fast transmission speed, and strong response ability.

Abstract:In this study, a 3D motion capture method is proposed for a flexible object with point features. Firstly, the method utilizes two calibrated high-speed cameras to capture motion video of a flexible object, and makes a stereo rectification of the image. Secondly, through the Difference Of Gaussian (DOG) algorithm, the positions of the feature points are obtained, and the extreme of feature points is extracted. Thirdly, matching pairs in a certain range of windows are searched, the feature points of the left and right images are matched. Thirdly, 3D reconstruction is realized by triangulation. Finally, using the search strategy to match the time series, the 3D motion capture of a dynamic flexible object is realized, and the spatial coordinates, velocity, and acceleration parameters are calculated. The experimental results show that the method is more accurate than the method of using SIFT algorithm to match the feature points to capture the flexible moving object.

Abstract:In order to overcome the shortcomings of the traditional security protection strategy based on the vulnerability database, the recognition and early warning of unknown attack behavior should be realized. Using time window division and deep packet inspection, the content of end-to-end communication is transformed into a sequence of control actions. According to the control protocol's semantic features, the control behavior sequences are transformed into the feature vectors of unified dimension using the semantic vector model. The anomaly recognition model based on One Class Support Vector Machine (OCSVM) is constructed by normal behavior samples only, overcoming the difficulty of obtaining exception samples from the production environment. The average recognition accuracy of the model is to more than 93% on the simulation sequences containing multiple abnormal behaviors.

Abstract:This paper introduces an approach to detect the quality defects of passports. The scumming defects of the passports are generally linear and produced in the course of passport printing. This phenomenon will have a severe impact on the overall quality of the passport. The chief difficulty to detect above defect is that the color of the scumming is close to the background texture of the image. Consequently, there is no relevant effectively detecting method. In order to detect the scumming defect, we first down-sample the image and analyze the multi-scale for reducing the influence of background pattern, and then employ a line segment detector to get these linear defects. The same line defect detection is conducted in both the samples and the template. We match the detected lines in each sample to eliminate the genuine lines while the non-matched lines are identified as defects. We also use the prior knowledge that linear defects are nearly vertical to improve the detection accuracy. Experiments show that the method of detecting linear defects is resultful. We deem that it is beneficial to control the passport quality and reduce the impact on the follow-up production of passports with this approach.

Abstract:With the coming of New Media Era, the spread of network information has become more extensive. Many online video evaluation activities were carried out by some institutions, while an evaluation method with rationality and fastness is lacked. In this study, we establish the hierarchial analysis model of many index signs on the basis of analyzing factors of online evaluation. The major steps of AHP and further PageRank optimization method are provided, which can be used to calculate weight values of index and obtain the final ratings. These research methods have certain reference values to dealing with multipleindexes appraisement.

Abstract:In the field of NPC game or GIS system development, there are more studies on finding the efficient method of the shortest path search problem, especially the research of A* algorithm efficiency optimization in the search of finding shortest path algorithm. Most of the maps used in artificial intelligence or algorithm research are based on arbitrary graphs rather than grid-based graphs. Based on above mentioned scenario, by combining arbitrary graphs with grid graphs and directions, an optimization of A* algorithm is proposed in this study. The heuristic search strategy is improved, which can reduce the scale and range of algorithm search and improve the efficiency of A* algorithm. Finally, the experimental results show that compared with the traditional A* algorithm, after the optimization of A* algorithm, the heuristic search strategy is more accurate and the computational efficiency became more efficient and more quickly.

Abstract:In order to solve the requirements of navigation robot for universality and compatibility, this study designs a navigation robot experiment platform accord to the concept of modular design, which consists of a microprocessor module, motor module, communication module, sensing module, and power module. Meanwhile, aiming at the cumulative error problem of the navigation robot in the construction of an environment map, a maximum expected probability mapping algorithm, which combines prior estimation and maximum likelihood estimation, is proposed to solve the cumulative error. Finally, the navigation robot platform is used to carry out the experiment of mapping and autonomous navigation. The coordinates obtained by the indoor motion capture system are used as a reference to validate the algorithm, which can effectively reduce the cumulative error generated by the robot.

Abstract:The diversity of Smart Grid services and the different QoS requirements are the urgent problems to be solved for allocating resources of power communication network. Network virtualization is the key technology of network transformation, which has a great advantage in ensuring QoS and improving resource utilization. Based on network virtualization technology, this paper describes the problem of QoS-driven resource allocation for power communication network and proposes a two-stage resource allocation model based on tripartite game, and then QoS-Driven Utility Maximizing Resource Allocation Mechanism (QDUMRAM) is proposed. It is proved that the QDUMRAM satisfies the compatibility of the dominant strategy and can achieve the goal of maximizing the system profit. Simulation results show that the QDUMRAM can maximize the utility and improve the resource utilization rate for the power communication network.

Abstract:Analyzing petroleum data from multi-source is a laborious and complex process, which often results in semantic and syntax conflicts. In this study, we take advantage of the knowledge representation and automatic reasoning of the ontology and propose an ontology-based multi-source petroleum information fusion framework. On the basis of this framework, ontology-based element similarity algorithm and fusion rules are proposed. Experimental results demonstrate that such framework improves the analysis efficiency of petroleum multi-source data.