最新刊期
卷 7 ， 期 5 ， 2002
- 摘要：This is the seventh in the survey series of the yearly bibliographies on image engineering in China. The purpose of this survey work is mainly to capture the up to date development of image engineering in China, to supply a convenient means of literature searching facility for readers working in related areas, and to provide a useful reference for the editors of journals and potential authors of papers. Considering the wide distribution of related publications in China, more than 400 image engineering research and technique references are selected carefully from near 2300 research papers published in a set of 15 Chinese journals. These 15 journals are considered as important journals in which papers concerning image engineering are relatively concentrated. Those selected references are classified first into 5 categories and then into 21 classes according to their contents. Some analysis and discussions about the statistics made on the classification results are also presented. This work shows a general and off the shelf picture of the various progresses of image engineering in China. In 2001, the worth well notified are the more attention on image formation techniques, the newly arisen research hot points related to intellectual property protection, multi sensor fusion and high level semantic information, as well as the broaden of application domains of image techniques.4320|449|0更新时间：2024-05-07
- 摘要：Optical flow estimation is a low level phase of motion recovery problem. Its output is a 2D projection of the velocity field on the image plane, which is analyzed further to get high level motion descriptions. A number of approaches have been developed for the optical flow estimation. The gradient based approach for the optical flow estimation is one of the importance kinds, it assumes brightness conservation under motion. This provides one constraint for the two unknown local components of the velocity vector. It is ill determined system of equation. A color image has some significant advantages over black and white image: it can give more optical information. Therefore, the colors images may be improve the solution of the optical flow estimation problem.The paper makes an introductory review on the optical flow estimation including the last achievement in this field. First, give the general concepts and fundamental principles of optical flow. Then, some methods for optical flow estimation are listed, black and white image and color image. Finally, make a discussion on some open problems and point out possible directions for further research.3173|359|0更新时间：2024-05-07
- 摘要：SFS(Shape from Shading) is one of the class ill posed problems of Computer Vision. General SFS algorithm often involved some additional constraints, such as smoothness constraint, to find a unique solution. However, this is based on the assumption that the reconstructed surface is continuous. Concerning the disadvantage of tradition shape from shading algorithm, a new algorithm of shape from shading based on fractal constraint is proposed in this paper. The method first introduces a novel fractal constraint, then through formulating the SFS problem using techniques of linearization of the reflectance map and minimizaiton of an energy function calculates the surface height. The proposed method overcome the disadvantage of tradition shape from shading algorithm, that is, distortion of recovered result of the 3D surface causing from assumed smoothness constraint, and does not require any integrability constraint or artificial assumption on boundary conditions. Experimental results show that new method is better than tradition method in aspect of 3D surface reconstruction for natural scenery.3249|365|0更新时间：2024-05-07
- 摘要：In computer vision, it is called multi views 3D reconstruction for recovering both camera and object shapes from multiple images, and it is currently a topic of lively interest. A hierarchical reconstruction method had introduced in 1996. In the course of completing the hierarchical reconstruction, the projective reconstruction is the first key step, having very important effect on the precision of euclidean reconstruction. The existence methods are very efficient for simulation data, but they are not perfect for real image. Namely, they are not robust and the reliable results can only be obtained if images match accurately. In this paper, the projective reconstruction based on genetic algorithms is proposed, the projective depths are coded by using decimal system and the adaptability function is defined by a constraint of the measurement matrix rank 4. The projective depths are iteratively estimated by genetic algorithms and Singular value decomposition (SVD) so that the measurement matrix is made to be as close as possible to rank 4, and then the projective reconstruction is realized. The validity and robusticity of the proposed algorithm is confirmed by experiments.2654|385|0更新时间：2024-05-07
- 摘要：Registration of serial images plays an important role in clinical research, diagnosis therapy and evaluation in medicine. A novel method used for serial image registration is proposed in this paper, which is based on the joint histogram of the two images to be registered. Usually, it is very easy, simply with a properly selected threshold, to segment human tissue from the background in medical images. With the thresholds of the two image sets to be registered, the joint histogram is divided into four separate regions. The configuration of the joint histogram bears a greatly different appearance between the before registered and after registered images. This difference appearance reflects the degree of the registration. The criterion function is defined as counting in a specific region or several regions of the joint histogram, which simplifies the computation complexity greatly and speeds up the alignment process significantly. With the fine property of the criterion function, powell's direction set method in multi dimensions is chosen to carry out the optimizing process to calculate the registration parameters. The comparison of the results from both mutual information and our method shows that the new method based on segmentation and counting is a fast, simple, efficient and accurate registration method.2834|348|0更新时间：2024-05-07
- 摘要：In 1968, Aristid Lindenmayer introduced a formalism for simulating the development of plant, subsequently named L systems. This formalism was closely related to abstract automata and formal languages. But there is a shortage in programming by the L system codes to construct plants graphic .The shortage is it's complicate and the low productive in its code because the program has to reuse the same code and traverse or modify the same linked list's each node to draw the plant graphic in every status. To make up this defect, this paper analyses and testifys a quality that those L system creating graphics in different status are of fractal character. Furthermore, the paper advances an improvement method based on this quality, which means that generation of the plant graphic in the later status can directly make use of the plant graphic in the former status. In other words, the later status graphic only is the former status graphic's enlargement?translation and revolving. Concretely, the new method is let the L system code into a simple recursion expression which can make the program code cut down and bring up the speed of draw a plant's graphic. A example of the new method is given at the end of the paper. So we can acquire a very useful conclusion in simulating plant with the L system theory which turn the L system code into a recursion expression.3088|403|0更新时间：2024-05-07
- 摘要：Image processing has to deal with many information of an image. Gray histogram can contain a lot of image information. Maximum entropy theorem of Information Theory is one of the useful tools to treat with this kind of information. There are several formulas for computing the maximum entropy. But almost of the existing formulas have some common disadvantages, such as expensively computing and more complex algorithm realizing. In order to overcome these weaknesses of the existing entropy formulas, in this paper we define a new approach to entropy, and use it to automatically select thresholds of the image. It bases on one of Shannon entropy's basic properties that the equivalent probability distributing has maximum entropy to get the image thresholds. And by this way, we can segment an image into several equivalent probability sub parts. This new method has some advantages, such as simplified, stabilized and easily realized comparing with some traditional entropy methods. At the same time, it can get image thresholds quickly. We have employed the newly proposed approach to perform image enhancement, segmentation and thresholding, and obtained satisfactory results.关键词：maximum entropy;threshold;Grey level;histogram3279|397|0更新时间：2024-05-07
- 摘要：Spectroface is a face representation method using wavelet transform and Fourier transform and have been prove to be invariant to translation and tolerant to expression variety. In this paper, the two important issues on Spectroface system is studied. One is how to preprocess system, another is the selection of similarity measurement. The moment is employed to preprocess system that it is good method to normalizing the scale and rotation of human face. The similarity measurement has been selected by comparing four typical kinds of similarity measurement, i.e., nearest neighbor method, averaging method, Hausdorff distance method and modified Hausdorff distance method(MH). Nearest neighbor method, averaging method and modified Hausdorff distance method are good for Spectroface. Nearest neighbor method is the most effective method in the recognition of frontal faces with translation, scale, rotation, different facial expressions, small pose, small occlusion and different illumination condition. It gives high accuracy as 97% and 99% in Yale and Olivetti face image databases respectively.2830|354|0更新时间：2024-05-07
- 摘要：This paper presents an effective license plate location algorithm, which employs color and texture analysis to extract the number plate from the complicated background image. Based on the HSV color model, the algorithm calculates the distance and the similarity in the color space to segment the color image. Only those parts of the input image that fulfill a set of license plate properties need to be considered for a more thorough inspection. To the segmented image, the texture and structural features are analyzed to locate the license plate correctly. The algorithm has been tested with 60 color image obtained from tollgete, crossroad and parking lot, etc. More than 95% image can be proceesed correctly, whereas other images are disturbed by the complex backgrounds of the images. It is shown that, different from most license plate location methods, the algorithm has fewer limits to the car size, the car position in the image and the image background. Meanwhile, with the fast focusing technique, the view angle of the video camera can cover a wider area while the processing time can still be fast. So, the algorithm can be employed to the applied vehicle license plate recognition system.3732|495|0更新时间：2024-05-07
- 摘要：The detection of transitions between shots in video programs is an important first step in analyzing video content. The wipe is a frequently used transitional form between shots. Wipe detection is more involved than the detection of abrupt and other gradual transitions because a wipe may take various patterns. This paper presents a new method for detecting wipe transition region and identifying wipe pattern based on pattern templates. In the proposed method, the dynamic pattern template(DPT) that stores the variance of the video frames in the wipe region is designed. Here, a special wipe video that only include black and white pixels are used to form the DPT. The DPT commonly has N frames. If the DPT satisfy the limit of one time scan, it can be simplifed to one frame. Then we give an algorithm based on DPTs to detect wipe region and identify wipe pattern. Here, only I and P frames are extracted from the MPEG stream for computing the wipe strip of two frames. Finally, to eliminate the disturbance of motion, Hough transform is used. Results show that the method can accurately detect wipe transition region and identify wipe pattern. Above all, the method is capable of satisfying the future wipe patterns.2886|363|0更新时间：2024-05-07
- 摘要：To realize ECG (electrocardiogram) and VCGr (reconstructed vectorcardiogram) combined diagnose, a solution is put forward to recover cardio signals from scanned standard synchronous 12 lead ECG image directly in order to reduce fund devotion. The cardio signals are complex, diverse and sensitive to noises, the characteristics of 12 lead ECG signals are similar with each other, and the qualities of scanned image are variant. All of these factors lead ECG signals difficult to be identified. To resolve the problems, original color image is firstly transformed to black-white image according to the defined threshold section. Then the least distance rule is applied to search the corresponding pixels for leads one by one in adjacent fields. After searching all pixels in the black white image, some missed pixels are compensated by interpolation. So the standard synchronous 12 lead ECG base lines can be recognized and ascertained. At the end, a filter is applied to remove gross errors produced by previous steps. Experiments show that the algorithm is robust and efficient because it can identify and keep the shape of original signal even when the signal contains some disease information or is corrupted by noises.2721|318|0更新时间：2024-05-07
- 摘要：According to the basic problem of the sub domains decomposition of simple polygon in the computational geometry and the finite element mesh generation of computational mechanics, an algorithm for decomposing a simple polygon into set of triangles and convex quadrilaterals is proposed in this paper in which rules of decomposition are developed. And these rules are propitious to the finite element mesh generation. The algorithm is guided by a weighting function for pairs of visible vertexes of the given polygon and the estimation factor of shape quality. The created weighting functions are related with the concave (convex) angles, so the function values are different for any two different concave (convex) angles. The algorithm is composed of three steps: (1) Decomposing the simple polygon into set of convex polygons step by step, (2) Decomposing the over six convex polygons into set of under six ones (include six ones), (3). Decomposing the six and five convex polygons into set of triangles and convex quadrilaterals.2950|357|0更新时间：2024-05-07
- 摘要：Typhoon is a kind of atmosphere swirl which spins round its center and at the same time moves along rapidly. The center collects most of the energy and the water and is the place most destroy happens. Thus its location is very important for typhoon path prediction and disaster forecasting. Typhoon center location task is the continuous research of "eighth five year project" presided by China Meteorological Administration. The tradition prediction method is poor of precision and its error is too great to meet practical requirement. According to the satellite cloud digital image feature and typhoon movement rule, researchers co operate with experts in shanghai central broadcast bureau and advance a new method. In this paper, the rotation matching method is mainly discussed. The method is based on the fact that although the typhoon mass is a non rigid object, the center area can be treated as rigid object and thus the center is one point with zero rotating vector. Thus in typhoon's sequence images, one can find the matching feature points and calculate the rotation center point which has the character of zero rotation vector. This method improves the precision and objectivity. To moving objects like typhoon cyclone, location method combining its moving feature is the certain reasonable way.3689|388|0更新时间：2024-05-07
- 摘要：Delaunay triangulation is widely applied in manifold fields and has a number of application dependent approaches. This paper briefly introduces its significant properties and popular generation algorithms. Then an empirically efficient algorithm for Delaunay triangulation using a uniform grid in 2D is introduced. This method first preprocesses the data, divides the whole point distributed area into grids with around the same number of grids and points, puts all points into corresponding grids. It begins with forms an initial triangle. While looking for connecting triangles, it puts all new edges of found triangles into a queue and remove the edges which are used by two triangles or are known as boundary edges out from the queue. Repeat the process until the queue is empty, then the triangulation is finished. This paper also shows the validity of the algorithm. The algorithm is very easy for implementation and uses computer resources of time and space more reasonably. Through tests with random generated data, its running speed proves fast and exhibits linear time complexity. It can meet the requirement of stochastic clustered dot screens in publishing, printing and dyeing systems.2693|390|0更新时间：2024-05-07
- 摘要：To improve encoding effectively of fractal vector quantization of image, after using quad tree to segmentation image adaptive, a non plane approximate method of image blocks Based on 3 D projection rule in orthogonal base is proposed in this paper. Furthermore, a novel algorithm for fractal vector quantization of still images is formed. The coarse image is constructed by the DPCM coding method of the projection parameters and the codebook of D value image coding is formed, and we use multiple resolution codebook when D value image coding. This method connects fractal and vector quantization coding together, during the process of decoding, the only need to check the codebook and to transform the contrast. It solves the shortcomings such as slow speed and lower compression ratio existed in traditional methods of fractal image compression well. The result of Computer simulation test of encoding and decoding show that the method does not need training to obtain the codebook in encoding, and not need iteration in decoding, so the method improves compression ratio and reconstructed image quality(PSNR) obviously.2859|348|0更新时间：2024-05-07
- 摘要：Digital elevation model(DEM) are collections of elevation points and could be used to describe the geographical features and to show the terrain surface in three dimensions over a certain area. This features are more important for various applications. DEMs are created by collecting elevations and referencing them to corresponding points in the mapped area. The elevations add a Z value to the ground's X and Y horizontal coordinates. However, it is always more difficult to get DEM and normally needs more computing time to generate DEM since its complicated algorithm and huge data. In order to speed up the generating of DEM, this paper makes some sensitive study on the parallel processing of the interpolation of DEM. Based on the analyzing of several common interpolation methods of DEM, the linear interpolation (sometimes called Proportional Distance Method) was chosen as our basic algorithm. Meanwhile, data parallel and algorithm parallel approaches were applied to this basic algorithm. Through the experiences at networked parallel processing system some good results were obtained and proposal for further research on this topic was provided.3272|350|0更新时间：2024-05-07
- 摘要：The paper, by means of illustrations, demonstrates the necessity of marking definitely design data in drafts, because these design data are the key factors of the embodying the intentions of the designer and the functions of components and parts. In combination with what may occur in practice, the author, as well, puts forward a set of approach that can be used conveniently without complicating the drafts .The approach is used to mark directly design data symbols consisted of signs and codes on the design data elements in drafts and the solution to the of marking of the design data symbols in computer aided drawing. After marking the design data symbols(two or three the symbols only) in drafts, the draft users will know where the design data are at once. Therefore, all kinds of the problems, which arise from the design data not having been marked in drafts, will be solved entirely. Meanwhile it points out when the design data symbols are marked directly on the design data elements of components and parts in drafts, other various problems like the conflicts of many kinds of dimensional tolerances coexisting in a same draft, the disagreement between noted dimensional tolerances and unnoted dimensional tolerances in a same one will be solved effectively.3165|362|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