最新刊期
卷 7 ， 期 8 ， 2002
- 摘要：A metamorphosis or a morphing is the process of continuously transformation one object into another, which is widely studied in graphics, simulation, animation and virtual reality technologies etc., and has great signification. Many metamorphosis techniques have been proposed in recent years for 2D or 3D objects. An almost comprehensive survey of the different approaches of object metamorphosis is presented in this paper. Study and development of metamorphosis problem is given in the field of the metamorphosis principle, 2D metamorphosis and 3D metamorphosis technologies, especially the technologies based on mathematical morphology. The final section analyzes the key points and the prospects of object metamorphosis, which lead to the important problems to be considered and solved.3311|365|0更新时间：2024-05-07
- 摘要：Motion capture has been one of the most promising technologies in character animation in recent years. However, most currently available motion capture approaches suffer from costly equipment and motion restriction caused by markers. To overcome these problems, a novel approach to extract 3D motion from video using vision technology is presented. The key issues, such as feature tracking and 3D reconstruction, are deeply studied. A model-based feature racking algorithm, which utilizes Kalman filter to predict coordinates of image features and epipolar line equation to aid tracking, is presented to track human motion with great variety. A non-coplanar nonlinear calibration model and a reconstruction approach taken uncertainty into consideration are applied to restore 3D human skeleton model. At last the experimental results and analysis of our VBHA (Video Based Human Animation) system demonstrate the feasibility and effectiveness of our approaches.2767|328|0更新时间：2024-05-07
- 摘要：In the new MPEG-4 video coding standard, the semi-automatic video segmentation plays a key role in supporting object-oriented coding and enabling content-based functionalities. A novel hierarchy optical flow based semi-automatic spatial-temporal video segmentation method is presented in this paper. The proposed algorithm comprises of spatial and temporal segmentation modules. In the spatial segmentation stage, the user can input points around the video object(VO) with the proposed point-based graphic user interface(PBGUI), then active contour model and tracking bug algorithm are used to precisely define the video object of interest to be segmented. With the result of spatial segmentation, the temporal segmentation involves non-rigid object boundary tracking and rigid object whole entity tracking by hierarchy optical flow algorithm based on the algorithm proposed by Lucas and Kanade. And the tracking points selection algorithm is proposed to greatly improve the tracking performance in the rigid object whole entity tracking. The experimental results show that the proposed algorithm can precisely segment video objects from video clips and can be applied to object-oriented coding, content-based functionality and multimedia database indexing.2471|314|0更新时间：2024-05-07
- 摘要：Taking into consideration the two clustering factors, the mutual distance between clustering objects and the centralizing effects of the higher level objects on the lower, a new clustering method based on minimum cost span tree with control vertexes is proposed. The MST is built based on the power of the clustering objects' mutual distance, and the selecting standard of the splitted edges is controlled by the higher level vertexes. Each splitted edge should be the longest edge under the condition that the two descendant trees must include at least one controlling vertex, and each descendant tree would include one and only one controlling vertex by the end of the algorithm. It has been verified by clustering the data built by ourselves and the earthquake data that this method, with simple input and little intervention, can discover better the true law of data distribution in some cases. To fulfill the needs of data mining, the selecting standard of the controlling vertexes, the 'inconsistent edges' and the efficiency of the algorithm should be improved.2772|318|0更新时间：2024-05-07
- 摘要：This paper presents a new classification-based fusion scheme, which is used throughout this paper as a synonym of the "pixel-level fusion". It can be applied to merge low-resolution images and co-registered high-resolution images. The key of this trial is as follows: (1) Geometrical co-registration IKONOS 1-m panchromatic image and 4-m multi-spectral image; (2) Classification of the High-resolution Panchromatic image together with the low-resolution images using supervised or unsupervised algorithms; (3) According to each class histogram of High-resolution image, adjustment of the corresponding spectral values of single multi-spectral image; (4) HIS transformation of the adjusted multi-spectral image using cylindrical coordinates, and acquirement of the fusion images. From this algorithm, we can conclude as follows: (1) The proposed method can merge the class information, spatial information of high resolution image and spectral information of Low-resolution image, accounting for the importance of classification as prior knowledge ; (2) Based on precise classification, it can effectively eliminate the false color at the edge of objective and have better visual effects; (3) It can try to add classification and spatial relationship to image fusion. In this trial the selected Peking image of IKONOS will be used, which can bring us convenience when verifying our algorithm.关键词：Image fusion;Classification;HIS transform;Prior knowledge;Image processing2645|332|0更新时间：2024-05-07
- 摘要：It discusses a new method of the description about the luminous intensity distribution of spotlight. An image with 256-brightness-grades on a plane which is perpendicular to spotlight's main axis is created, from which the lighting direction and intensity can be calculated by each pixel's brightness and location. By this way, the luminous intensity data of luminaries or lights can be accessed, stored or calculated easily, and the expression for unsymmetry-luminous-distributed spotlights is more accurate. This method may take an important role in the simulation and visualization of lighting environment. Using a picture of brightness to express the LID of the spotlight is correspond to put infinite curves from all the connected planes in range of 360 degree surrounding the main light axis to a picture and display it directly. It is especially convenient to the LID-adjustable spotlight used for TV and movie. A system is developed using a set of camera and white diffuse reflection wall to take the lighting picture from which the luminous intensity distribution image can be created after being processed according to the camera exposure latitude. Example pictures of lighting scenes using the distribution images are given.2632|275|0更新时间：2024-05-07
- 摘要：This paper put forward a method of image edge detection and filtration on the curve coordinate system by the use of operator on the curve coordinate system.Edge detection is a important process in image processing, currently the popular method firstly to be used is the Laplacian operator on the Cartesian coordinate system, then the zero-crossing of the Laplacian operator can be found as edge point; image filtration is the convolution result of the differential operator on the Cartesian coordinate system too. It is not fit to deal with many images painted on surfaces such as cylindrical map image, spherical map image,digital effects for motion pictures and medical regime on three dimensional scans of organs. We make use of the differential operator on the curve coordinate system for processing images on surfaces. The practical numerical experiment results showed that differential operator on the curve coordinate system is capability for processing images on surfaces.2521|333|0更新时间：2024-05-07
- 摘要：Mismatching points are inevitable in scene matching because of the complexity of the matching conditions. In order to improve the reliability of scene matching guidance system, it is necessary to develop the methods of automatic recognition and elimination of the mismatching points. Due to the limitations of the aircraft maneuverability, the flight paths are rather smooth, and sometimes are linear, especially during the process of scene matching. So the relationships among the correct matching points in one matching area are regular. Three algorithms are developed according to the changing regulation of the fitting curve of matching points. The three algorithms are as follows: (1) on the basis of small curvature of the fitting curve of the matching points; (2) on the basis of the linear curve generated by the matching points; (3) taking advantage of the accuracy of the distance-measurement accuracy of INS in short period of time. These three algorithms are tested by the data gained in flight-test, and the results show that, all of the three algorithms can distinguish and eliminate the mismatching points effectively and can meet the need of real-time ability of the matching system, though the real-time ability, adaptability and reliability are somewhat different. The proper algorithm should be selected according to the practice conditions in flight, and algorithms (1) and (3) or algorithms (2) and (3) can be fused according to the matching conditions to improve the eliminating trustability of the mismatching points.2819|342|0更新时间：2024-05-07
- 摘要：Essentially most commonly-used denoising methods use low pass filter to get rid of the noise. But both edge and noise information are high frequency information, so the loss of edge information is evident and inevitable in the denoising process. Edge information is the most important high frequency information of an image. Therefore we should try to maintain more edge information in the process of denoising. Thus comes out the idea of this paper. We present a new image denoising method:wavelet image threshold denoising based on edge detection. Before denoising, those wavelet coefficients of an image that are corresponding to image's edges are first detected by the method of wavelet edge detection. The detected wavelet coefficients will be protected from denoising and therefore we can set the denoising thresholds only based on the noise variances without damaging the image's edges. The theoretical analysis and experimental results presented in this paper show that, compared with the commonly-used wavelet threshold denoising methods, our denoising method can keep image's edges from damaging and increase PSNR up to 1~2dB. Finally we can draw the conclusion:Edge detection and denoising are two important branches of image processing. If we combine edge detection with denoising, we can overcome the shortcoming of the commonly-used denoising methods and do denoising without blurring the edge notably.2950|337|0更新时间：2024-05-07
- 摘要：Moving and stationary target acquisition and recognition(MSTAR) program has shown that segment synthetic aperture radar(SAR)imagery into taeget,shadow and background clutter regions is a efficient measure in the process of recognition targets in open terrian.But traditional image segmentation methods are unable to achieve precise segmentation owing to the image affected by speckle noise.In this paper, SAR imagery segmentation algorithm based on MRF(Markov random field) is proposed. The prior information about the segmentation image with MRF model is presented, the prior probability distribution of every region is got from training data by maximum likelihood(ML) estimation,the Bayes formulation is adopted to obtain the conditional distribution of the posterior distribution of the segmentation image conditioned on observed image,based on the maximum a posterior(MAP)criterion,the segmentation is abtained by Metroplis algorithm.By applying this algorithm to the MSTAR sample target images,the result demonstrates the algorithm can achieve robust and precise segmentation result.3022|371|0更新时间：2024-05-07
- 摘要：Color image segmentation is essentially a clustering process in 3D color space, and color images can be considered as s special case of multi-spectral images. It is quite often when the objects cannot be extracted using three features but can be extracted using more than three features. In this paper, every pixel of an image is described using 5 features which are transformed from its own RGB features, then multi-dimensional thresholding(MDT) divide color space by thresholding each component histogram. This approach is equivalent to partitioning the multi-dimensional histogram into rectangular hyper-prisms. But MDT will lead oversegmentation in result. Hence, two-step algorithm is proposed to solve the problem of oversegmentation. The first step is region growing in 3D histogram by connectivity of frequency of neighbor bin in same class. The second step is data clustering based on scale space theory, which models the blurring effect of lateral retinal interconnerctions. The final algorithm that combines different approaches results in further improvement in performance. It is simple and effective to different color image, and has been used to segment the microscope medical image successfully.2669|383|0更新时间：2024-05-07
- 摘要：This paper presents a new algorithm for reconstruction of a 3-dimensional object using silhouettes subtracted from 2-dimensional photos. The basic idea is to shoot a series of photos around a 3-dimensional object from different angles, subtract the boundaries of the object from the photos, and then combine the 2-dimensional information by tracing the rays to obtain the 3-dimensional coordinates of the points on the approximation surface of the object. Compare to some available method based on the concept of "voxel", our algorithm has the following advantages:(1)It significantly saves space because only the 2|dimentional mask boundaries need to be saved in memory; (2)The accuracy of approximation is independent of the 3-dimensional resolution since we are not using voxels; and 3. the speed is also greatly improved especially when the object has large volume since instead of projecting every voxel, only the points on the surface are computed according to our method.2563|345|0更新时间：2024-05-07
- 摘要：An improved OSEM(ordered subset expectation maximization) reconstruction algorithm with OR(overrelaxation) parameter is studied in order to provide a more rapid and practical convergent iterative reconstruction algorithm in tomograph of nuclear medicine, such as SPECT(single photon emission computed tomograph) and PET(positron emission tomograph). Based on the additive version of OSEM, the new method introduces a constant overrelaxation parameter z>1 during each sub-iteration, and imposes the non-negativity constrain and total-counts normalization condition in subset iteration to overcome the negative image value problem and total-counts shift, in order to ensure its rapid and stable convergence. The method is reduced into ordinary OSEM when z=1. The reconstructed images were compared with those of standard OSEM, by both simulated phantom data and clinical SPECT myocardial perfusion data. The OR-OSEM is shown to be one time faster than OSEM's with same subset level, and is even faster than the OSEM with higher subset level. The results also show that this OR-OSEM method is more flexible in practice due to its continuous adjustable OR parameter.3505|366|0更新时间：2024-05-07
- 摘要：Face modeling is an important step in model based face reconstruction. A method adapting the facial wireframe model from two face images(front one and side one) with some user's interaction is proposed in this paper. At first the place of the front face and the feature regions in the front face image are located by region growing and template matching,the deformable templates are used to extract the full facial features. Secondly the exact front positions of features are rectified by hand with a friendly interface, the depth positions of the feature points are defined from side image manually. Finally the rotation of the head is calculated and the model is scaled and the other vertexes of the model are adapted with inverse distance interpolation algorithm with the feature points as the data points and then the input face model is gotten. The test results show that this method is simple and useful.2653|379|0更新时间：2024-05-07
- 摘要：This paper introduces a method to recover seriously deformed palm-print images via three-channel wavelet transform. Those filters sharing common input or output signals with other are called as a filters bank. The deformation processing in this paper is just like digital resample, in which a filters bank with three filters is implemented to the palm-print images for wavelet decomposition, and then a procedure of binary interpolating is performed after the image is reconstructed by another filters bank which also has three filters. A way on QMF method to design multi-channel wavelet filter banks is introduced in this paper, too. Because the wavelet transform has the feature of frequency division, the images are de-noised, eliminated the sample effect(the Morie stripe), at the same time. The 3-channel wavelet transform extends multi-resolution wavelet analysis that can only be based on the scales 2j, thus makes the multi-resolution wavelet analysis more in detail. The method is feasible and the results are acceptable.2570|321|0更新时间：2024-05-07
- 摘要：In order to meet the instant need of individual identification,the whole project of an automatic fingerprint identification system based on oriented filtering and segmentation is presented. The project is mainly divided into two parts that are fingerprint image preprocessing and fingerprint recognition. It includes calculation and smoothness of directional image,oriented filtering,binarization,noise cleaning,feature extraction,feature coding and matching. By the methods of preprocessing,feature coding and fingerprint matching presented in this paper,the fingerprint image is effectively enhanced,moreover, during the course of fingerprint matching,certain influence of fingerprint rotation and moving can be avoided. The system has a high rate of fingerprint recognition and a strong ability to resist image noises. In addition,it is proved to be high effective in recognizing damaged or blurry fingerprint images. The system is high valuable in practice.3004|340|0更新时间：2024-05-07
- 摘要：The issue of car license plate recognition is a focus direction of studying both at home and abroad at present, its success has important application values in vehicle controlling, transportation management, parking and so-so. There are many papers published in the research domain. In order to solve the primary problem of the car license plate recognition--Automatic orientation technology of the car license plates, a method of automatic detection and orientation of car license plate based on license plate's projection invariability in the condition of lesser deflection of car license is presented in terms of the imaging characteristics of the car license plate target, which can succeed in detecting and orienting the car license plate from the complicated background. Tested actually through the scene, the method obtains satisfactory localization effect. In the end of the paper, some experimental results are given out. In virtue of the successful car license plate detection and localization, it is possible for license plate number extraction and recognition.2641|351|0更新时间：2024-05-07
- 摘要：The electron gun of the picture tube is very importance in TV, the characteristic of the bend and distortion affect the quality of picture tube, so we have to measure the electron gun's bend and distortion error. According to the electron gun's structure characteristic, a new project about the measurement of the electron gun's bend and distortion error is designed. This paper introduces the image processing technique that contain Auto-adapted two-value method,image filter, laplacian edge detection, reformative hough transformation and their application in the measurement of electron gun's bend and distortion error. The experimental results show that these methods can effectively increase the measuring precision and the measuring speed of the electron gun's bend and distortion characteristic.2241|309|0更新时间：2024-05-07
- 摘要：Firstly, the development and application of digital city are introduced. Aiming at the complexity of constructing digital city, the key issues and implementation scenario of constructing digital campus based on GIS component are explored, which will give a practical method and useful reference for digital city. Adopting the GIS platform Geo-Union, a digital campus of Peking university (PKU campus) is built. The architecture of PKU campus is based on Browser/Server mode, which consists of four layers: Browser, Web Server, Geo-Union Server and Spatial database server. The spatial information of PKU campus is organized as spatial entity, map layer and map and stored in Oracle8i.According the distribution of the campus, the spatial information consists of five parts: environment, main campus, Yan-bei campus, Chang-ping campus and medicine campus. The functions of PKU campus include information services both inside PKU and around PKU, such as map exploring, Web linking, site locating, spatial querying and navigating. The implementation method of PKU campus is based on WebGIS component and ASP. Two secondary developing components (GisView and PicView) are designed to integrate visualization, map operation, human-user interaction for map and image.PKU campus integrates all-sided service information in PKU, which will provide a convenient information channel for all members of PKU and other users, and will be a new window for exhibiting the image of PKU.2802|343|0更新时间：2024-05-07
- 摘要：In order to realize Geo-information sharing, the first step is to share data, moreover, semantic sharing is its second and higher object. Semantic interoperability in geographic information system, which is the same as application level interoperability, is one of the available approaches to realize Geo-information semantic sharing. In this paper, status quo of interoperability and semantic interoperability in geographic information system are firstly probed into, and some existing methods to realize semantic interoperability are reviewed. Then, the new road—metadata mediator has been presented. The metadata mediator consists of semantic knowledge and rule bases, a semantic conflict finder and a coordinator. The mediator is implemented in the three-tier client/server architecture. Here, formalizing the metadata of geographic data and semantic expression of geographic information based on the formalizing method are two key tasks. By this mediator, semantic conflict can be recognized and processed, and semantic information mining also can be implemented. Furthermore, A prototype of metadata semantic interoperability, which is built according to the metadata mediator theory, is discussed, so far the prototype has shown its power in the National Resources, Environment &District Economy Information System (NREDIS).2440|307|0更新时间：2024-05-07
- Postal code：100190
- Tel：010-58887035/58887030/58887418 Email：jig@aircas.ac.cn
- Technical support is provided by Beijing Founder electronics co., LTD 京ICP备05080539号-4
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
0