最新刊期
卷 7 ， 期 3 ， 2002
- 摘要：Wavelet image denoising has been well acknowledged as an important method of image denoising. Based on many literatures of wavelet denoising, this paper attempts to make an overview of wavelet image denoising. First, it describes wavelet denoising in two ways, one from its mathematics background, the other from filter theory of signal processing. Then this paper classifies wavelet image denoising methods into three classes that includes shrinkage based method, projection based method, and correlation based method, and describes them respectively. Considering the important role that coefficient model plays in a wavelet based denoising scheme, this paper also discusses three kinds of wavelet coefficient model, including intra level model, inter level model, and hybrid model that combine the first two together. Usage of simultaneous wavelet transformations, the relationship between wavelet image denoising and wavelet image compression and wavelet denoising under different noise models are also covered here in order to give an overview as complete as possible. At the end, the future trend of wavelet image denoising is pointed out, though, in personal opinion.5324|623|0更新时间：2024-05-08
- 摘要：Image interpretation is an important part of computer vision, which is related to many fields such as image processing, classifier designing and logic reasoning. In this paper, genetic searching based two directional reasoning is discussed. The algorithm consists of two steps. At first, the fuzzy memberships of classification is obtained by fuzzy classifier based on the statistic/geometric features of segmented regions, and a fuzzy graph used for effectively representing interpretation information is constructed through prior rule base concerning about spatial relations that is acquired from statistics or experience. At second, genetic searching algorithm is used to combine the above two types of information, and the optimistic interpretation is achieved. In order to decrease the computational cost and increase the possibility of getting optimal solution, a new crossover operator of genetic searching is proposed that is based on non random graph partition. The experiments show that genetic searching based fuzzy image interpretation is powerful for the regions that include one or more objects. This method is an improvement over the one directional reasoning method based image interpretation such as probabilistic, evidence and fuzzy reasonings.2385|433|0更新时间：2024-05-08
- 摘要：Volume rendering 3D imaging is a new technique for visualizing 3D arrays of sampled data. It has wide applications in areas such as medical imaging and scientific visualization. But due to vast computing quantity of volumetric data sets, its application has been limited by its high computational expense. At present, many researchers are working hard to the foundation of accelerating algorithms. In order to resolve the rendering speed and imaging quality problem in 3D Visualization of medical image, this paper present a interactive dynamic volume rendering algorithm, which is a rendering method from arbitrary viewing point distance and by arbitrary viewing ray direction, and then, basing on analyzing its algorithm complicacy, proposes a accelerating algorithm. It apply the six features of 3D nonuniform transformation, and use the simple beeline marching and parallelogram marching instead of time consuming matrix multiplication. This makes the volume rendering reach to almost real time efficiency. This method is four to five times faster than the standard one.2814|428|0更新时间：2024-05-08
- 摘要：The traditional graphics--surface graphics only can display the surface of the 3D objects.It can not display interior properties of the 3D objects, such as texture, density field and temperature field.So many fields such as computer graphics, CAD and finite element analysis, need a new visualization technology for displaying the interior of 3D objects. The visualization technology of 3D object is an important part of a new graphics--volume graphics. Lots of information of solids, such as texture and density, are often hidden in the interior of 3D Objects. Based on the volume graphic theory and the characteristic of octree, this paper proposes a technology for interior visualization of the 3D objects for the demand of CAD system.The octree arithmetic is applied for voxlization which converts geometric objects into a set of voxels that best approximates the continuous object.As a result of uneven changes of interior properties,the technology uses the unequal voxels to best represent the details of interior properties.According to many practices, the technology not only can represent interior properties sufficiently,but also has some modeling functions.2690|520|0更新时间：2024-05-08
- 摘要：The clustering phenomenon often appears in the cell image auto reading system, i.e., some cells overlap or touch together to form a big area. It is necessary to design an effective algorithm to separate the clustering cells into single ones. Most of the existed separating algorithms are based on two hypotheses: 1) at points where cells touch, the cluster boundary tends to form an acute angle; 2) at points where the cells touch, the optical density is relatively low, otherwise it will be very difficult to separate them properly. But in the actual cell images, the two hypotheses are very difficult to be met. To solve this problem, an algorithm based on boundary stripped is presented in the paper. The boundary of clustering cells is stripped layer by layer in the algorithm. During stripping, judgement is made to determine whether the splitting has happened, and then the actual separation is taken in the original cell image. It does not require the clustering cells should necessarily meet the two hypotheses. The experiment result shows that the algorithm is effective.2517|460|0更新时间：2024-05-08
- 摘要：In order to robustly and accurately restore the 3D vision information of an object from its two perspective views, by means of the new idea of randomly sampling the minimal redundant subset, by utilizing the data regularization technique, we develop a new algorithm, which can robustly and accurately recovers the 3D vision information of an object from its two perspective view data--the set of their feature point pairs. Random sampling can significantly reduce the sampling number of subset and make the good subset surely sampled. The redundant information contained in the minimal redundant subset can be efficiently used to check the validity and goodness of the sampled subset. The data regularization technique can greatly alleviate the numerical unstability generated from the ill posed property of the data. So, the algorithm is able to work well with high accuracy under very hard condition of heavy noise and high outlier rate. The experiments have demonstrated that the processed results are satisfactory.2206|396|0更新时间：2024-05-08
- 摘要：The weakness of the conventional box counting algorithm in measuring the dimensions of fractals is analyzed. A novel improved box counting method is proposed by introducing a new parameterη, which is the ratio of fractal object's length in a box over the edge length of the box. For fractals having infinite self-similar levels,η is also infinite which has no meaning in mathematics. But the fractals existed in the real world all have their scaling limits, soηis meaningful in practical measurement. Its advantage is proved by the examples in the paper, and its limits are also pointed out. The uncertainty in measuring the fractal dimension using box counting method is discussed, and the corresponding means to overcome them are also proposed. Planar laser-induced fluorescence technique was employed to measure the concentration field in the near wake region of circular cylinder. Turbulence non-turbulence interfaces were extracted from these images, and the new box counting algorithm is used to measure the fractal dimension of these interfaces. The measured results agree with those of other researchers, and it is found that the fractal dimensions vary little with time in an unsteady flow field.2877|441|0更新时间：2024-05-08
-
Small Target Detection Based on Energy Accumulation and Order Morphology Filtering in Infrared Image
摘要：Because of the influence of nature meteorological condition, background environment and the structures of objectors, detection of weak and small objectors in infrared image is one of difficulties among image objector detection and identification. A new small target detection method based on energy accumulation and order morphology filtering is presented for solving detection of moving small target with low SNR in infrared image sequence. Accumulating the energy of the image sequence in the sliding window with a setting size can remove the random noise in order to increase SNR. Three steps of target detection from coarse to fine are adopted. First suppressing background and extracting generalized edge of targets by using order morphology filtering, determining the prime location of the targets, then segmenting the regions where there are possible targets, extracting the geometrical features. Finally filtrating the true targets according to the principle of moving continuity and trajectory consistency of moving target in the image sequences. The experiments show that this method can effectively suppress background and extract generalized edge of targets, adaptively choose threshold for segment, perform the location and detection of small targets in infrared image.2400|304|0更新时间：2024-05-08 - 摘要：Morphing technique of human face image is a key technique in the field of face modelization. Based on geometrical feature of facial contour, this paper selects appropriate Feature Points and constructs a component feature mesh model adapted to face image morphing through the muscle movement principle of face organs. In order to imitate the muscle movement of facial component, such as eye, nose, brow, mouth and face contour, feature mesh constitutes two kind of synthetic muscle. One is the contour muscle which manipulates the removal of vertical direction. Another is the linear muscle that controls the displacement of horizontal direction. Thus, the movement of muscle point can be divided into two directions, vertical and horizontal, which is controlled respectively by the contour muscle and linear muscle where the point locates. Feature mesh model ensures that facial component morphing is manipulated in virtue of the change of component contour, coinciding with psychological cognition. This makes it more adapted to human image morphing than other morphing algorithms. The result of experiment shows that the facial images which generated by this algorithm are accurate, and the expression effects are more natural.3403|570|0更新时间：2024-05-08
- 摘要：Multi band wavelet is a newly developed branch of wavelet analysis and is superior to 2 band wavelet in terms of symmetry, compact support and smoothness. There have been many studies on multi band wavelet's theory and applications in the recent years. In this paper, based on multi band wavelet transformation and interpolation theory, a new approach to magnify images is introduced. Firstly, we analyze the lack of the methods for the image magnification by general methods such as 2 band wavelet method and linear interpolation. Then we discuss the basic theory and analyze the transformation properties of multi band wavelet. Next, we study the image magnification methods by the double linear interpolation and the weight parabola interpolation. Further, we propose an approach that magnifies an image based on multi band wavelet transformation and interpolation theory. Especially the three band wavelet transformation is implemented to magnify real remotely sensing image. Furthermore, the magnification results by multi band wavelet are compared with other results based on commonly used interpolation magnification methods. Finally, the magnification images are evaluated quantitatively by mean gradient. The theory and experiments show that magnification images by multi band wavelet can retain the characteristics of original images very well and are superior to the magnified images by other commonly used interpolation methods.2647|428|0更新时间：2024-05-08
- 摘要：Visualization of oil reservoir model is one of the most important steps of reservoir description. From the two or three dimensional images of oil distribution provided by the visualization result, oil engineers will make more exact decision. At the beginning of this paper, the procedure of visualizing the reservoir model is introduced. After that the vital problems are analyzed. Oil water borderlines, top model and bottom model are all constructed to determine the location and shape of the reservoir; then physical property field is constructed by using Kriging interpolation method. In the end of this paper, by implementing all the techniques refereed, an example is given. The result shows that, by successfully implementing all the techniques refereed, the reservoir model has been displayed vividly in three dimension space.2297|392|0更新时间：2024-05-08
- 摘要：Getting the curve border from a black white image is very important in CAS/CAD/CAM. A new algorithm for it is presented. The main idea of this algorithm is auto segment of edge points and quick Bezier fitting to those points. In this algorithm, edge points of a black white image are detected firstly. At second step, these edge points are segmented according to left and right direction of every point. Then, these segments are classified into two kinds: curve segment and line segment. Those points on the curve segments will be fitted into Bezier curves using a certain simple and quick method. In this way, a set of border from a black white image, which consists of lines and Bezier curves, is presented. Comparing to the vector method, this algorithm can get accurate and smooth border of a black white image which is not distorted even if it is enlarged many times, so it is easy to output the border of a black white image to devices such as sign maker, plotter, CNC machine, etc. This algorithm is efficient, and widely used in serial software "Wentai Kehui", the speed and result of which are as good as the same kind of software made in developed countries.3180|621|0更新时间：2024-05-08
- 摘要：Data quality has been the vital factor on which the reliability of the result of GIS analysis and application depends with the development of GIS. Based on the review of GIS database quality control recently research, a multi hierarchy conceptual model of data quality control is given in this paper. We can accomplish data quality control beginning with data source and then database proceeding, finally data using by this model, and then aiming at the characteristics of Marine Fishery GIS data, the detail structure of database quality control is schemed, at the same time several practical approaches of data quality control are discussed thoroughly. At last, based on the techniques of metadata and data warehouse, a prototype of intelligent data quality control is provided.2609|521|0更新时间：2024-05-08
- 摘要：In this paper, after an analysis and comparison of data structure between MapInfo and AutoCAD, first a method for distilling elevation points, elevation values and other correlative information from graphical symbols and labels in CAD by matching area and transforming types is put forward. Second a means is designed to create building polygons by the way of recurring topological relations between points and lines, on the basis of lines that represent buildings in CAD. Finally, automatically acquiring spatial and attributive features of terrain elevation and buildings is implemented. Making use of these processed data, an urban DEM model and a display usage of 3D images are constructed. Getting cross the procedure of data processing in which data transform from CAD format to GIS vector data then GIS raster data and 3D images, an integrated processing method of 3D urban modeling is formed. The method adopted in this paper is more simple and faster than the ways and means of engineering air photogrammetry, and it has a more practical value. Here, an example based on the terrain data of the 32th Jiefang in Suzhou is given by using the integrated method of 3D urban modeling and a fairy good result is attained.2713|503|0更新时间：2024-05-08
- 摘要：In this paper the authors present a novel integer reversible DCT based embedded image coding scheme, which can integrate lossy and lossless coding schemes perfectly. Image coding is of great interest in applications where efficiency with respect to data storage or transmission bandwidth is sought. The discrete cosine transform(DCT) has been applied extensively to the area of image coding. It has nice decorrelation and excellent energy compaction properties and it can be easily implemented by hardware. In this paper the integer DCT transform is implemented by factoring the float DCT transform matrix into integer reversible transform series. Then the authors apply the series of matrices to image samples, and encode the coefficients by several effective coding algorithms. The simulation results illustrate that the integer DCT coding scheme is superior to the float DCT coding method in lossless coding, and the coding performance of integer DCT is close to that of float DCT in lossy coding.3260|551|0更新时间：2024-05-08
- 摘要：In this paper, a region segmentation based multiresolution motion compensation wavelet video coding algorithm is proposed. The features of the kernel parts of this coder are as follows: 1)In the part of motion estimation, based on the basic variable block multiresolution motion estimation algorithm ,we take full advantage of the spatial and orientation correlations between the motion structures represented by the wavelet decomposition subbands and improve the coding efficiency of the motion vectors; a region segmentation based multiresolution motion estimation algorithm is presented, under this algorithm, the computation complexity of motion estimation is reduced and the motion vector coding efficiency is improved farther. 2) In the part of quantization, a quantization scheme based on human vision system is proposed. 3) An improved version of zerotree wavelet coding algorithm with high performance is adopted to code the intraframe and interframes. The experimental results prove the effectiveness of the algorithm this paper presented.2280|448|0更新时间：2024-05-08
- 摘要：Because there may be bit error and packet loss in network transmission, the network between an real time MPEG 2 encoder and decoder would bring impairment to the bitstream especially when the network is heavily loaded. The video quality would be degraded when the bitstream is corrupt or broken. In this paper, a MPEG 2 TS stream rate control algorithm based on active packet discarding is proposed to resolve this problem. It test the network traffic by monitoring MPEG 2 stream transmission buffer, and discards some not so important TS packet of the MPEG 2 bitstream actively according to the test result to reduce bit rate before pouring data into network. When data loss is inevitable, the difference between active discarding and data loss in the network is that we can control which part of data should be lost under active discarding, and we con't know which part of data will be lost if we leave the loss control to network. Experiment results demonstrated that this algorithm is simple, useful and resultful for real time MPEG 2 TS stream transmission in network. It is adaptive to network traffic, the receiver end video quality is improved obviously, especially when the network is heavily loaded.2659|480|0更新时间：2024-05-08
- 摘要：In this paper a novel MPEG-2 transcoder based on DCT domain is proposed. After analyzing the differences between MPEG-2 MP@ML and 4∶2∶2 Profile as well as the properties of discrete cosine transform and some spatial domain filters, the upsampling and linear filter algorithm in DCT domain is derived. The new transcoder has low buffer requirement and results in low delay. Compared with common decoder-encoder cascade methods, the current algorithm can reduce the computation and hardware complexity dramatically due to its omission on DCT(IDCT), ME(motion estimation) and MC(motion compensation) procedures,which are very time consuming. The simulation results demonstrate,the new approach only suffers slighter quality degradation than the pixel domain method on basis of the conventional decode/re-encode schema, and it is more appropriate to be used in real time environment.2185|398|0更新时间：2024-05-08
- 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