The radar/IR integrated dualmodel seeker is widely applied in missile guidance system. The process of simplifying the target detection of IR imaging system with the relative distance between missile and target provided by radar detecting/tracking system is discussed. The calculation methods of detection probability and false probability of radar/IR integrated dualmodel seeker are also presented.

Fiber-optic guidance technology is one branch of precision guidance technologies which is in a flourishing development. The technical characters and basic principle of fiberoptic image guidance system are introduced. The key techniques of the system are discussed emphatically, such as, imaging seeker, fiberoptic bidirectional transmitting system and image tracker etc. An actual fiberoptic guidance system is cited. Furthermore, the application direction and development of fiberoptic image guidance techniques are analyzed.

Three image feature indices (edge density, roughness and standard Dvalue) of every pixel in reference image are taken for the quality index of estimation of template image matching. With Fisher linear classifier, a new approach to estimation of matching probability is proposed which transforms the calculation of matching probability into the classification of pixels. The utility of the method is demonstrated by experiment and suitable matching areas can be sifted out.

IR imaging guidance technology that adopts Focal Plane Array (FPA) has been widely used in airtoair missiles by reason of its predominant capability. On the surface of aircraft, different parts are at different temperature, so the different parts of target image are different in grayscale. It will cause difficulty in accurate target segmention. Based on the features of aircraft IR image and the purpose of IR image segmentation in imaging guidance technology for new airtoair missile, a novel segmentation algorithm that is beneficial to target features recognition is proposed. The target image is segmented firstly. Then the areas which are inside of the target image are filled by selecting the suitable structural elements and utilizing the parallel characteristics of mathematical morphology. Finally, the accurate result of segmentation is obtained. The segmentation result of this method is of great help to grayscale continuity based on feature recognition algorithm.

Edge detection plays an important role in lidar image procession. Its capability has direct influence on precision and performance of lidar imaging system. The principles and performance of five edge detection algorithms(Sobel operator,Kirsch operator,LOG operator,Canny operator and parameter verification operator)are analyzed and compared.Experimental research with the five operators is made on real lidar images with speckle noise. The results show that the parameter verification operator works better.

The SNR equation and detection probability of a direct detection laser radar are discussed. The software for modeling the signal of lidar echo is developed based on MonteCarlo method. The simulation results of a typical pulse lidar show that the horizontal detectable range reaches 5 km with the pulse energy of 1 mJ under the visibility of 5 km.

The development and actuality of laser passive reconnaissance warning device are sketched. The composition, operation principle and features of the system are introduced. The design of laser probe and data processor of the laser warning device are emphasized. InGaAs Pin photodiode is adopted in the optoelectronic detector of the probe. The data processor comprises threshold circuit, discrimination circuit of pulse bandwidth and twoecho signal restraining circuit. Firstly, the laser signal is digitized under the condition that the signal is affirmed. Then the incident orientation of laser is calculated. Finally, the laser ranging signal and guiding signal are distinguished by threat style and grade curcuit on the basis of threat database. The system has the advantages of high sensitivity, rapid response and wide dynamic range, and it can be linked up with smoke discharging equipment. The system and other defense equipments can be connected by preparatory communication interface.

One 2 kW high power diode laser was used for cladding of high speed steel powder on the substrate of a tool steel. A crackfree cladding layer could be obtained along the interacted zone of the focused powder stream and focused laser beam. With the increasing of laser power, the thickness of cladding and availability of powder were increased, and the dilution of substrate to cladding layer was decreased. The hardness of the cladding layer was up to 800 Hv0.3 and the hardness of the substrate was up to 200 Hv0.3. The result demonstrates that high power diode laser is very competitive for future application in surface cladding

According to the theory of orthogonal experimental method, some factors affecting IR radiation features of tank are analyzed by computer simulation tests with the theoratical model of IR radiation feature and the software for tank. The main factors affecting IR features of tank are summarized. The monofactor analysis of the main factors is made. It is significant to the application and verification of the theoratical model.

Y1Ba2Cu3O7-xsemiconductive thin film with different buffers is manufactured by current magnetic sputtering with direct and annealing method, and their IR thermal radialization responses are studied. The microstructure of Y1Ba2Cu3O7-x thin film is analyzed by XRD and Raman spectrum. The electronic resistance temperature coefficient value and hall effect of Y1Ba2Cu3O7-x film are measured. The results show that the uniformity of the semiconductive Y1Ba2Cu3O7-x thin films with buffer layer is better and their signaltonoise ratio is higher. Y1Ba2Cu3O7-x has good speciality for its thermal radiation response in IR band and has good performance in submillimeter band, even in millimeter band. It will become new sensor elements of IR bolometer working at room temperature

The characteristics of small targets, noise and background in IR image are analysed. Because the small target is often corrupted by high noise after high pass filters, a peakvalue detection algorithm with wavelet transformation is presented basesd on the softthreshold denoising method proposed by Donoho. The simulation results show that the proposed algorithm can effectively restrain the residual high noise filtered with high pass filter and detect the small target

A novel system to independently detect small moving object in IR image sequences taken from a moving platform is proposed. Egomotion effects are removed through a waveletbased robust multiscale affine image registration process. The low frequency images are segmented with the improved Fisher algorithm and wavelet transformation, then the object is separated from background. The difference of the segmented image after registration is computed, and the kinematic track of the object is obtained. The experimental result shows that the system is effective.

Feature points extraction is crucial for image segmentation and moving object modeling, and uniform sampling always causes lost of details. Illumination variation and noise also make the extracted feature points unreliable because of the sensitivity of classical detector to noise. A method based on 2D Gabor wavelet is presented to extract features, and a dynamic feature clustering based on feature GWT energy is used to compute the local feature points. The method is robust to noise and the features points extracted are very close to true local features. Active mesh based on image features can be generated by feature points extration.

Based on the characteristics of the frequency spectrum of remote sensing images, an adaptive layered quantization fast DCT image compression method is proposed.The JPEG quantization table is adaptively modified after the original image is discrete cosine transformed, then the DCT coefficients are layered quantized using the new quantization table. The real remote sensing image compression experiments show that the proposed method runs faster than standard JPEG compression method at the same compression rate, and its peak signaltonoiseratio gain is 1～2 dB. Moreover it can be used to realize embedded code stream image compression.

An embedding strategy and an examination algorithm for digital watermark based on wavelet transformation domain are proposed. Digital watermark is embedded into the middle frequency subband images of the original image in discrete wavelet transformation domain. The experimental results show that the embedded digital watermarks of the proposed algorithm are invisible and robust enough against the generally used image procession techniques such as Guassian, Pepper, and JPEG procession etc.

License plate location is essential in license plate recognition. A algorithm for license plate location based on adaptive signal peaks detection and projection is put forward in view of the shortcoming of general projection method. Firstly, the input image was segmented horizontally by projection. Then vertical projection was used to obtain potential regions with license. In the end, false regions were deleted. Experimental results prove that the algorithm is effective.

Computer based 3D face model has been widely used in many fields, such as virtual reality, computer aided instruction, teleconference, HCI, computer game, movie making, etc. The method of Dvalue is presented to realize the generation of personalized 3D face model by modificating a generic 3D face model only from front view and side view face images. On the basis of the generated personalized 3D face model, face animation system based on MPEG4 framework is realized. The experimental result is better.

A great quantity of information in the face image makes it improper to be used to direct recognition. Discrete cosine transformation is used to process raw face image, and dimension reduced feature vectors are obtained. Feedforward neural netwok is also used to classify and recognize face. Simulation experiments are conducted based on face images in ORL face database. The results show that the recognition rate is quite high and the train time is notably shortened, so the method is efficient for face recognition.

The technology of modeling and simulation for dynamic IR image is discussed. Due to the complexity of the target, background and atmosphere, the difficulty for the modeling and simulation of infrared image is great. The design idea, achievable way and software frame for the modeling and simulation software of IR image sequences in SGI workstation are introduced. The imaging under various radiation influence is analyzed and the calculation formula of radiation is provided. In addition, the passive scene and active scene are analyzed separately. Then the calculation methods of temperature in passive scene are provided, and the functions of the scene model, atmospheric transmission model and physical property of materials acting on IR imaging are explained. Finally several examples of the modeling and simulation for infrared image sequences as well as the synthetic estimation are presented.

Radar imaging geometry, which performs imaging in the direction of slanting range at the sample time of electromagnetic echo reflection, is quite different from optical imagery. The hypsography results in not only the layover distortion of groundtruth under different imaging conditions, but also the inconsistency of intensity distribution of imaging gray. Consequently the positioning performance of radar scene matching is directly reduced. A real time adjustement method based on vertical projection, by means of analyzing the relationship between the ground elevation and the model of radar imaging, is proposed to adjust the geometric distortion of sidelooking real aperture radar image. Experimental results show that the method is effective.

The principle and application purpose of oversamplesuperposition technique are discussed based on the characteristics of long wave IR detector. The effects of oversample and superposition are analyzed on the basis of PTF, and the corresponding oversamplesuperposition PTFs at different division multiples (m) are given. As m is increased, the pattern of pushingtype scan sample is varied from instantaneous single sample (m=1) to integral sample at full residence time (m is tended towards infinity). The variation of PTF is not distinct when m>4.

The methods of locating the high light area of image, determining the color value on the edge of this high light area and the color rectification are discussed to resolve the problem of color saturation of traffic sign from crossing. ExpectationMaximum of GMM based on symbols is used in MaximumLikelihood Estimation and good classification result is gained. Then a fast algorithm for computing region depth is used to mark the depth of highlight area. A color rectification model is proposed and the experimental results and discussion on the methods are given.

Based on the properties of infrared radiation windows and harmonic diffractive element, harmonic diffractive element is successfully introduced to the system of infrared dual color detector. The rectification of the longitudinal aberration in two bands is simultaneously accomplished. Wavefront aberration is less than 1/4 wavelength and the modulation transfer function of dual band approaches or attains the diffraction limit. The results show that the properties of action spectrum of harmonic diffractive element are between refractive element and diffractive element. The system debases the demands to technical level. The practical design shows that the system not only has the advantages of compacted structure, few elements and high rates of transmission, but also has better weaken ray aberrations character. It can be used to put forward a new element for infrared optical designs.

Design and implementation of a realtime transport cascade multiplexer in SDTV are discussed. The main devices of the system are DSP and FPGA. It can multiplex multiple SDTV programs into one TS stream through cascade of multiple multiplexers. Setting up a higher code rate, many null packets in TS stream can be generated by upper multiplexer, so local generated TS packets can be inserted into location of these null packets by cascade multiplexer. Through multiplexer cascade, multiple SDTV programs can be multiplexed into one TS stream.

DSP has been widely used in various information systems and interintegrated circuitl I2C bus is a popular communication bus used to communicate between different chips. Therefore it is very important to effectively design an interface between DSP and I2C. Through the deep analysis of the writing and reading time diagrams of DSP and I2C bus, a solution (hardware and software design) to implement I2C interface based on DSP is proposed. By use of the external memory interface of DSP,fliplatch and buffer, DSP can be communicated with I2C devices correctly. The writing and reading flow of I2C devices are controlled by software and high reliability is achieved through simple hardware devices. The example of TMS320C80 DSP is also proposed. 