In order to reduce the complexity of the algorithm of image enhancement, to simplify hardwareimplementation and to get more clear and natural image, we propose a new histogram statistics method, Presentsan improved histogram statistics and distribution adjustment algorithm. On the whole distribution histogrammodification the brightness saturation problem is improved. Uses the fixed point of histogram segmentation pointsarea balanced histogram. With histogram after balanced reflects the enhanced image. Experimental results showthat the method in the case of adding less amount of calculation to ensure the image has a better effect, not onlyensure the oversaturation did not appear, and guarantee the image detail is outstanding, has a good visual effect.The enhancement effect of this method is superior to the traditional method.

In order to achieve self-management and improve the personalization and intelligence, the Smart-homeplatform architecture is presented in three layers, which is named as resource layer, adaptation layer andapplication layer with perception ability. Based on the atom Web services for the household-equipmentmanipulation, a smart-home services SOA (Service-Oriented Architecture) framework is proposed by binding thegiven Web services. Experimental results show that it could unify perception and control mode for equipmentfrom different smart-home alliances from different networks, it can build collaborative platform for intelligentservices developers and home equipment suppliers, reduce the difficulty of the two sides to interact with data, in order to better develop smart home services in their own areas.

This paper is based on the DCT(Discrete Cosine Transform) basic principle and characteristics, andits purpose is to break the shackles of the unified characteristics of traditional DCT transform matrix, for theoptimization on the DCT transformation matrix of image information, to achieve the image compression efficiently.Firstly, compare the reconstructed image by DCT with the original image. Then, use genetic algorithm to find theoptimal solution by minimizing the mean square error, in order to optimize the DCT transform matrix coefficient.Finally, use the transform kernel optimized for image processing. Experiment was carried out from blocking theimage and setting genetic initial parameters, and the experimental results indicate that for the linear edge andtexture feature images of small block processing, when the number of population is 40 ~ 60, the length of thestring is 8, the crossover probability is 0. 7 ~ 0. 8, and the mutation probability is 0. 007 ~ 0. 008,the DCToptimization method based on genetic algorithm for image compression to reach the best effects.

Different from ordinary facial expression, micro-expression has the characteristics of short duration andlow intensity, which is often difficult to be identified effectively, resulting in the limitation of research in this field. Dealing with these difficulties, a novel feature combination method was proposed. Calculation using the global optical flow technology was carried out in the adjacent frames to obtain faint optical flow. The motion information of each two adjacent frames was passed by, so that the changes were significant between the images of two frames at a distance, solving the problem of short duration and weak action. The optical flow feature and space-temporal local texture feature extracted by LBP-TOP(Local Binary Patterns from Three Orthogonal Planes) operator were combined to make supplement for describing most region details of the faces. The random forest classifier was selected, and experimental results show that the two features have a very good complementary, in the CASMEII database. The method is able to identify five types of emotion, the accuracy is promoted from 40. 50% to 64. 46%, and the category discrimination is improved accordingly.

In order to control the temperature in the heat exchange system, the two degrees of freedom PID(Proportional-Integral-Derivative) control algorithm was used. The PID controller is synthesized by the Ziegler-Nichols oscillation method, and the coefficients of two degrees of freedom controller were adjusted according tothe empirical formula. A series of simulation experiments were carried out with the reference of the traditionalPID controller, including object tracking, anti-jamming, robustness testing and nominal model controlling. Thetwo degrees of freedom controller showed good performance, object tracking (better than the traditional one),and robustness. In the nominal model control simulation experiment, the system with the two degrees of freedomcontroller worked with a less rise time and a less recoil than which with the traditional controller.

In order to improve the performance of the BFA(Bacterial Foraging Algorithm), Bacterial foragingalgorithm and immune algorithm were combined, the clonal selection ideas in immune algorithm were used toreplace the reproduction operation of BFA. For the chemotaxis operation, the moving step of bacteria is shortenby the iteration proceeding, so that the astringency is guaranteed, and the overall searching capability of bacteriais ensured. The elimination and dispersal operation is improved by guaranteeing the bacterial with the highestfitting value not be dispelled to increase the astirngency accuracy. The results show that the optimal value whichwas obtained by the authors were closer to the optimal value than BFA's, which proved the algorithm was morecapable in optimization. Moreover, the algorithm was more stable because the variance of three functions were all less than BFA's.

The water damage problem is more and more worse in recent years, so the problem of algae image  classification and retrieval is extremely urgent. To solve this problem the algorithm based on content is proposed. SIFT(Scale-invariant feature transform) algorithm is used to extract shape feature of algae image because marine biological image is color insensitive. The dimension of the feature vector generated by SIFT can reach up to 128. The higher dimensions can affect classification prediction and can lead to high computational complexity. PCA(Principal Component Analysis) technology is adopted to reduce the dimension of the feature avoiding dimension disaster. K-means is employed to clustering algorithm, which is simple and effective. The result of the clustering algorithm is packaged by Bag of Words in order to subsequent identification. KNN (K-Nearest Neighbors) algorithm is used to recognition. The experimental results are finally consistent with the fact. It provides effective support for the research of environmental problems which is caused by algae.

In order to solve the complex object shelter in the scene and poor robustness, we put forward the algorithm based on template matching with linear prediction. The algorithm includes two parts, template matching and linear prediction. When there no obstructions in the scene, the executive part of template matches algorithm for geting target tracking information; otherwise, through the process of prediction algorithm to estimate the states of the target, and then to realize tracking. The results show that the method can track target accurately and stably for the target to keep out interference, comparing with the traditional template matching algorithm it has strong robustness.

In order to solve the current life-limited problem of the LED(Light Emitting Diode) with the power increasing, we made an optimization design of the existing LED track light radiator structure and the driving circuit. This design chooses the graphical interface application Ansys as a development platform, integrates
design of the simulation combined with the experimental data by using Gambit subdivision and Fluent software temperature field simulation. We choose the 50 W power LED heat distribution model as an example, present a distribution mechanism of LED temperature, the analytical model of the temperature field, the theoretical formula, the experimental field measurement data and the analysis of the simulation graph. The research showed that the new LED optimized thermal design realized the increase of LED power from 30 W to 50 W, while achieving the LED using temperature decreased from 80 ℃to 76 ℃. So the model achieved the extension of service life and the improved effect of lighting.

In order to solve the problem of only a single energy source existed in portable energy devices, we designed a set of portable energy device based on photovoltaic power generation and piezoelectric power generation technology, which has two main energy sources, solar energy and kinetic energy produced by exercising. The solar energy and kinetic energy are harvested through solar cells and piezoelectric materials, respectively. We designed the layout structure of the energy device to meet the demand for the portability and high efficiency. According to the result of the research, the energy device having the advantages of portability and high efficiency meets the technical requirements of portable energy devices, and can satisfy the practical needs in daily life.

Traditional K-eans algorithm of the initial clustering center is randomly generated, which can lead to roduce very big volatility clustering results. In order to solve this problem, We propose a algorithm named lustering algorithm based on min-istance Product. With the method of sampling, CAMDP(Clustering Algorithm ased on Min-istance Product) produces selected point which has minimum product of distances between itself nd all other initialized clustering centers, which improves the selecting of the initial value of the K-eans lgorithm, avoiding the random selected clustering centers. The results show that the topological feature is onsidered and the attributes of vertex are taken into account, which let the improved K-eans provide the strong upport to the division of community.

In order to improve the signal to noise ratio and fidelity of seismic data by the compressed sensing reconstruction method, an algorithm of compressed sensing reconstruction of seismic data based on curvelet transform is proposed. A seismic data reconstruction model is built, the energy and entropy distribution characteristics of multi-scales seismic data based on sparse representation of curvelet are analyzed, the computational complexity of the random observation is reduced with the block compressed sensing technology, with the change of information entropy, an adapt bivariate shrinkage threshold iterative reconstruction method is designed based on the correlation between the high frequency region of multi-scales in curvelet domain. Experimental results show that the proposed algorithm gains above 1. 5 dB, and has better ability to maintain the detail information than the other algorithms mentioned, under the same sampling rate.

In order to achieve remote sensing digital image fast displaying, by a case study in ERDAS IMAGINE IMG standard data format image file, we gave environment configuration process, development frameworks, established processes and realization of functional modules of the program of IMG image display and coordinated real-time display under the Windows system. The program realizes displaying IMG and real-time coordinate by use of the open source GDAL(Geospatial Data Abstraction Librarylibrary), a graphical user interface framework QT as a development tool, and Visual Studio 2008 as the development platform. The results show that Multi-threaded sub-block processing technology shortens 2. 7 s for 1. 8 GByte remote sensing image compared to the use of pure RasterIO (), improves the efficiency of image reading and display picture and meets the large amount of data applications.

In order to analyze the social group network visually, and to make accurate observation of the socialgroup behavior, the social network analysis method named social network data visualization is presented. Thesocial relation network was described as composed of point and line chart. Moreover, we make effective analysisto the node distribution location, the size of the graphics and dotted line density. Using the online visualizationanalysis technology for social network data, combined with the python visualization and python scientificcomputing technology, the system enhances the adsorption of websites, and it has gained a good user experience

In order to improve the test speed of the traditional serial cluster level, we study the parallel method ofcluster test tasks. In this process we through the dependencies between the classes and between the class tests tomake the cluster level test task parallelization problem transformed into solving the parallel nodes in objectrelational graph. First, complex cluster relationships could be formed into the loop. In order to determine thestarting position of the test, we need to break loop. We put forward the broken loop method considering the nodedegree, number of loops and complexity of test stub of the three kinds of factors. Second, by means of analyzingthe elimination of the loop object relation graphs, a hierarchical parallel algorithm is designed. Finally, anexample of serial scheme and hierarchical parallel scheme comparison demonstrate that the hierarchical parallel scheme has a practical significance.

To satisfy the rapid development of mobile terminal and the high personalized demand of mobile phone user, a design of mobile phone recorder system based on Android was proposed. The system can monitor and analy of mobile phone users and records, to judge whether the user should reasonable arrange the using of mobile phone, and gives a reasonable proposal, which will be convenient for users to monitor and manage by themselves. It mainly uses Intent and log monitor to implement software monitor function, encryption of file system through the class FileEncrpter, and for recording encrypted stream encryption is used, the use of class MediaRecord and ACC(Advanced Audio Coding) code to realize the audio recording etc. The system has achieved satisfactory results, and proved the feasibility of the system through experiments.

A method based on incomplete binary decision tables for fault tree construction is proposed to solve the multi-state problem and the reduction of decision tables. This method starts from the listing of all the possible event sequences by means of decision tables with the condition stub of events and the condition entry of 0 and 1. The output events can be deduced according to interaction of parts under circumstances of different event sequences. On basis of the working principle, the irrelevant and repeated events can be ruled out so as to simplify the rule. And in way of achieving the union sets of the event sequences with the same output events, the causal relationship among events can be achieved. Verified by the fault tree modeling of a nose wheel steering system, the result achieved in this way is consistent with the one in FMEA(Failure Mode and Effects Analysis) and the fault diagnosis handbook and the scale of calculation can be reduced effectively.

In order to improve the use of multiplexer for combinational logic circuit design methods, the multiplexer and or formal logic expressions of the output function is analyzed. The relationship between output function and output inverse function is mutually inverse. For some design can make the data input variables of the multiplexer be the original, without the need for additional gate to take the operation of the variable, simplify the circuit structure.