We propose a novel and efficient hierarchical scheme, which can locate the accurate positions of the eyes from frontal face images. First, Gabor transform is used to calculate the salient map and a number of rectangular regions with the maximum saliency values are selected as the coarse eye region candidates for further verification. Second, the two eye windows with the minimum PCA(Principal Component Analysis) reconstruction errors among the eyecandidate regions are selected. Finally, the pupil centers are localized by applying two neighborhood operators within the eye windows. The proposed algorithm is non-iterative, computationally simple and robust to different facial expressions. Experimental results on JAFFE database show that this algorithm can make the detection accuracy of 99.6%, and can achieve a superior performance compared to other state of the art methods.

In order to meet the control performance requirements of ETCS(Electronic Throttle Control System), such as fast computational performance, low-cost, miniaturization, FPGA(Field Programmable Gate Array) is adopted as the hardware implementation of controller and integral separation is added to mature PI
D(Proportional, Integral, Differential) control algorithm. And the integral separation PID controller with Catapult C highlevel synthesis tools is implemented based on FPGA full hardware solution. An integral separation PID controller on a FPGA chip is designed for ETCS. In addition, AD/DA acquisition board is designed to implement the data transmission between integral separation PID controller based on FPGA and electronic throttle. Finally, tracking experiments of electronic throttle are performed and the results demonstrate the effectiveness of the designed integral separation PID controller based on FPGA.

For the defects that the image can only be decomposed into a finite orientation by the wavelet transformation, and multidirectional of image can not be better characterized, the directional filter is applied to the method of wavelet threshold, namely, the HWD(Hybrid WaveletDirectional filter banks) instead of the mere wavelet transform is used to better characterized the multi-directional of image and to retain more image information. To avoid the phenomenon of amplitude cross exists between image wavelet coefficients and noise wavelet coefficients, the method of wavelet thresholding was restricted by this shortcoming, membership function was changed based on the principle of the wavelet, and the membership weights of HWD was built by the membership function. The wavelet coefficients of noise were set to zero maximum with losing the image coefficients at least by this improved HWD. The wavelet coefficients of noise were set to zero maximum with losing the image coefficients at least by this improved HWD. The studies show that more details of engineering drawings can be retained by the improved wavelet threshold method beside removing some noise.

To realize precise control for CSTR(Continuous Stirring Tank Reactor) systems in actual reactions, a robust L₂L_∞ state feedback control problem is studied by using a T\|S fuzzy model to approximate the nonlinear object according to reaction characteristics of the CSTR. All reactor temperatures which are easier to be measured can be treated as premise variables in the model, where fewer number of fuzzy rules are employed. The design method of partial controllers is given by applying the PDC(Parallel Distributed Compensation) algorithm. A global controller is designed by adopting the LMI(Linear Matrix Inequality) technique, and the design of controllers is cast into a convex optimization problem. Simulations are provided to demonstrate the effectiveness of the proposed design scheme, which can be applied to other chemical industrial reactions.

To solve the problem of aerospace equipment miniaturization, we use photocell array structure to replace the slits, and combine the slit with the photocell.Changing the diffracted light into a secondary diffraction, improves the signal quality. Using photocell array structure, reduces the size and the effects of uneven light, conducive to industrial production and aerospace equipment miniaturization. Practice has proved, cancel the slit to using an array of photovoltaic cells viable, incremental photoelectric encoder is an important improvement.

To deal with the problems of time and pulse lager synchronization in the process of CCD (Charge Coupled Device) camera capturing laser spot images,
 an advanced forecasting method to trigger CCD camera synchronously to capture images of high-precision clock source is proposed. This device used the UTC(Universal Time Coordinated) and high-precision PPS (one-Pulse Per Second) as synchronous clock reference source provided by BD2 (BeiDou2 navigation satellite system)/GPS (Global Positioning System) dualmode receiving unit. Ensured the synchronization capture time of the CCD camera in combination with high-speed FPGA (Field Programmable Gate Array) computing and microprocessor interface timing technology, thus accomplished the synchronization trigger action. The experiment shows that this device can provide microsecond level synchronization accuracy and standard timing information, shorten the exposure time of the CCD camera effectively, get a complete clear pulse laser spot image with a higher SNR(Signal-to-Noise Ratio).

In order to monitor the CO gas in the environment effectively, a new CO sensor which could work at room temperature was designed and fabricated. The sensor is composed of active carbon filtrated cap, Pt sensing electrode, Nafion membrane, counter electrode, the isolation and reservoir. The design of the reservoir provides the humidity environment that Nafion membrane needed for the normal work. The sensing properties of the sensor are mainly determined by the Pt sensing electrode. The preparing conditions of chemical deposition were mainly investigated. When the concentration of H2PtCl6 and NaBH4 are 5 mmol/L and 30 mmol/L, respectively, and the pH value of NaBH4 solution is 13, the sensor gave the best sensing performance. The response current is 107 nA, and the response and recovery time are 40 s and 50 s, respectively, when the concentration of CO is 400×10^-6. When the CO volume fraction ranges from 50～400×10^-6, the response current value of the sensor has a good linear relationship with the concentration of CO gas.

Aim at the issue of vehicle power battery such as long charging time, the charging process fails and can not timely access to the results after charging, the vehicle power battery monitoring system is designed based on the internet of things with the voltage of power battery, current and temperature as the monitoring parameters. We analyze and describe the overall system and then complete the building and the testing of the hardware design and software plat
form. Test results show that the human-computer interface using the LabVIEW virtual instrument platform is able to perform the real-time remote monitoring. Monitoring subsystem and controlling the charging data transmission between the subsystems are effectively accomplished. The system real-time performance and accuracy are verified, providing conveniency for users.

In order to study systemically the construction of integrating operational circuit and the selection of parameters we deal with the design problems of the integrating operational circuit. Expressions of the output waveform amplitude for square wave input are derived from circuit equations and integral conditions. And the results provide the theoretical basis for the design of the integrating operational circuit. An example is presented to demonstrate the validity, simplicity and practicability of the proposed design method.

To solve the the metal/cubic boron nitride (c-BN) contact problem, we put forward the Si doping of cubic boron nitride single crystal on the surface of the Cuzinc alloy probe, the surface uses the Ag plasma sintering, the preparation of a space charge limited schottky diode. The experimental data showed that the characteristics of IV at room temperature, the device in DC10 V when the rectification ratio is 370; the positive current and voltage is two times the power function relationship. With the space charge limited current and thermionic emission theory, the experimental results are analyzed, and the equivalent circuit model of the sample device.

Finger vein image has a low contrast and polarization characteristics and traditional segmentation algorithms can not solve these problems. C-V model contains the global information of the image, it can detect weak edge information. This article improves based on the classical model of the C-V model, increases the distance penalty term and edge detection function for fast and accurate finger vein image segmentation, it also solves the problem of uneven distribution of gray which the classical CV model could not solve. The improved algorithm preserves the characteristics of global optimization C-V model while effectively avoiding the re-initialization process, and it is more sensitive to edge. The simulation results show that the improved C-V model is better able to deal with the problem of fuzzy boundaries and uneven gray, it also can improve the efficiency of segmentation.

According to the integration difficulties of the multi-sources mass isomerism and distribution 4G geological spatial data, the merits of three commonly used technologies of federated-database middleware and data-warehouse is demenstrated, geoscience data integration and management model based on data warehouse and middleware is proposed; with 10 minerals of baishan mining area in Jilin province in 2005 year as data sources, the SOA(ServiceOriented Architecture) is designed based on SQL(Structured Query Language) Server to verify. The results show that the architecture is suitable for heterogeneous distribution and network transmission condition, can satisfy the requirement of data update fast and real-time consistency condition, and is easy to implement on cloud platform.

In order to achieve the flexibility for relative cable network and the inexpensiveness for 3G/4G mobile network, contenting to the urgent needs of the wir
eless network coveraging to the enterprises and units, we carry out the comparison research, give the implement under different conditions for the WiFi(Wireless Fidelity) coverage. To improve the security of wireless coverage systems, we carry out a scheme for the wireless WiFi coverage and offer to the con
crete implementation method system based on Web mandatory safety certification.It provides theoretical reference and technical support for a wide range of
 wireless WiFi coverage, and offers practical, efficient and safe wireless coverage solutions, it also provides implement support for small business institutions.

The smart home terminal control system based on Android is designed for greatly satisfying the user's requirement to living, providing better control experience of smart household appliances to the users. The terminal control system has the features of low cost, flexible operation and friendly interface which are supported by all kinds resources of Android, it can realize the remote control to the smart household appliances.

Surface defect detection and size measurement of the toner cartridge is an important part to determine whether it is qualified. To achieve this aspect of work automation, the toner cartridge surface point defects and line defects are accurately detected and classified by mathematical morphology operators of image processing algorithms, and its size is precisely measured according to sub-pixel measurements. The cartridge automatic defect detection and size measurement system is developed by the special machine vision software Halcon and .NET. Testing shows that the system has a high accuracy, speed, stability and other distinctive characteristics.

In order to solve large\|scale integration of national forest resource management, a distributed spatial data storage system suitable for forestry resources data is designed. Through three steps, block data files, index classification and merge cluster, without changing the underlying database software structure, provinces and national's data clusters are logically combined into a whole big data cluster. Application of the distributed spatial data storage system and merging of clusters increase the speed of data processing, unifiy data management process, establish omnidirectional concurrency data channel, is an active exploration for realizing forestry resources data for national integration application.

In order to solve the problems of real\|time vehicle tracking and separation in the traffic test system, using digital image processing method, a moving vehicles detection algorithm based on transportation videos is proposed. For the obtained traffic test information, the research focuses on the background subtraction algorithm to extract moving vehicles. A method of adaptive background updating algorithm with smaller amount of calculation is proposed and a nonmodel-based approach working in HSV(Hue, Saturation, Valve) space is used to detect vehicle shadow. Through setting the proper threshold parame
ters, we can remove vehicle shadow and remove disturbances of pedestrian, bicycle and motorcycle vehicle at the same time. After the shadow detect
ion, some suitable morphological methods are used for post processing. Based on large number tests of real traffic environment video and images, the proposed method is proved to be effective for moving vehicles detection.

To solve the problem of the construction of continuing education information platform in hardware system, based on the analysis of the construction of continuing education information platform in application planning and the principles of constructing hardware systems, design and construction of information platform runtime environment, we proposed the construction scheme of hardware system of continuing education information platform. It can provide protection for the students online learning and school's information work.