最新刊期
卷 7 ， 期 2 ， 2002
- 摘要：Watermarking is one of the current copyright protection methods that have recently received considerable attention. In this paper, we introduce a wavelet transform based watermark casting scheme for digital images. The proposed method searches the significant coefficients across subbands to embed the watermarks into the significant wavelet coefficients. The watermark is adaptively weighted in different subbands according to the recommended mask values of wavelet coefficients by Watson. The current quantization threshold for each subband is also used as the weighting function of the embedded watermark. The fidelity of the watermarked image can be adjusted by using a weighting factor of the cast watermark. The casting and detecting of watermarks is hierarchical. In the case when the received image is not distorted significantly, the calculation of cross correlations with the whole size of the image may not be necessary, therefore much of the computational load can be saved. Experimental results show that the DWT based watermark approach can provide an excellent protection under various attacks.2703|324|0更新时间：2024-05-07
- 摘要：A number of supervised and unsupervised pattern recognition techniques have been proposed in recent years for the tissue segmentation and quantitative analysis of magnetic resonance images. However, the accuracy of these methods is affected seriously by the intensity inhomogeneities of images. In this paper, We present a novel algorithm(BAFCM) for fuzzy segmentations of images that are subject to intensity inhomogeneities, such as magnetic resonance image. The algorithm is formulated by modifying the objective function in the fuzzy c means algorithm to include a gain field, which models image intensity inhomogeneities. First and second order regulation terms in AFCM algorithm ensure that the gain field is both slowly varying and smooth, but increase complexity of computation greatly. Instead of computing gain field, we compute bias field first, then convert bias field to gain field. With BAFCM, we can correct the intensity inhomogeneities and implement fast classification of human brain tissue of MR image automatically.2585|478|0更新时间：2024-05-07
- 摘要：Fuzzy learning vector quantization (FLVQ) algorithm outperforms the hard-competitive vector quantization in that it reduces the dependence of the resulting codebook on the initial codebook selection, yet it has the disadvantages of slow convergence and easy to be trapped in local minima. In this paper, the principle of fuzzy learning vector quantization for image coding is reviewed. Followed by a discussion of the possible ways for optimizing the FLVQ algorithm, a new fuzzy learning vector quantization algorithm based on tabu search(TS-FLVQ) is then proposed. In this algorithm, we firstly constructed a table listing oriented to global search by the tabu search algorithm, and afterwards took advantage of fuzzy learning to reach the global minimum point of the predefined objective function. The algorithm with a detailed description of the procedure involved was simulated in the computer finally. The algorithm differs from a standard greedy search in that the best move is executed also if it leads to a configuration with a greater energy than the current one; this is necessary to be able to escape from local minima. Experimental results show that TS-FLVQ has much better coding performance over FLVQ with remarkably faster convergence.2675|606|0更新时间：2024-05-07
- 摘要：We aim at finding a suitable quantification algorithm to encode color invariant for indexing and retrieving images. To this end, an adaptive cluster network quantification algorithm for color invariants is proposed. By using a set of training images to train the color invariant vector, this algorithm can acquire a suitable color invariants vector having an adaptive number of members. In this paper, we discuss in detail how the threshold and the step in the algorithm influence the number of the vector members. After having done many experiments, we get a vector of 29 members for our training images when the threshold is 1 0 and the step is 0 3. In this setting, the vector is rather robust. Then we also apply the algorithm and another bench algorithm, named averaging quantification algorithm, to a content based image retrieval system. Experiments have been conducted on a database consisting of 1 126 images taken from different image databases. In order to evaluate and compare the querying results, an application specific software is developed. In the view of the correctness of the querying results, comparing the adaptive quantification algorithm with the averaging quantification algorithm, we find the former is superior to the latter by 4%. In the view of the time complexity, although the former takes a long time to train quantification vector and to acquire a lookup table for the image database, it is much superior to the latter when retrieval proceeds. Finally a conclusion is obviously obtained that adaptive quantificaion algorithm is an excellent quantificaion algorithm for color invariants.关键词：CBIR;Color invariants;Adaptive cluster network quantification;Cluster;Averaging quantification2514|322|0更新时间：2024-05-07
- 摘要：To realize the real time video coding in low bit rate communication by PC software is the hotspot in image compression field, thus a frequency domain block matching algorithm based on discrete cosine transform(DCT) coefficients is proposed in this paper. By utilizing threshold matrix, the motion estimation computation is greatly reduced and the time used in searching motion vectors is saved. The DCT based frequency domain block matching algorithm can encode and decode motion pictures quickly and efficiently. Its reconstructed picture quality is higher than the 3SS algorithm and is closer to the FS algorithm. Simulation results show that the proposed DCT based block matching algorithm and the video coding scheme has high performance and is easy for implementation. Because of its low computing complexity, it can be real time realized by PC software and has a good value in practice.2368|321|0更新时间：2024-05-07
- 摘要：This paper presents a method of image multi scale retrieval. That is by means of wavelet transform, which firstly decomposes an image in multi scale, which yields a scalable coefficient matrix; secondly determines the F norms of all the ordinal sub matrixes of the scalable coefficient matrix; and then defines the features and similarities of images by the F norms; finally uses these similarities index images. But in order to speed up the computation, this paper only deals with the Integer to Integer Wavelet Transform, the main advantages of which are: all the calculations can be done in place; the input and output values are all integers; and only a small memory is required. Additionally, the feature vector based on F norm is low in dimension, and efficient in retrieval. The experimental results of a database of about 200 color images of flower and mount show this algorithm can well capture the scale and color information of images and can lead to better results.2322|330|0更新时间：2024-05-07
- 摘要：In this paper, a new approach for automatic extracting the local maximum(or local minimum) lines from the first horizontal derivative images of geophysical fields is proposed. The basic procedure of this method is as follows: First, the directional profile method is applied to recognizing local maximum (or local minimum) pixels from the field images. Then, the rotation transformation method is applied to searching maximum (or minimum) line endpoints in all directions. Finally, a vector map for the local extreme lines of the first horizontal derivative images of geophysical fields is obtained through that, the closely spaced lines are merged into broken lines in line with the line merged regulations, and that all of short lines are eliminated. The aeromagnetic map of Shandong peninsula was served as the experimental image. The experimental results showed that the local extreme lines both extracted automatically and interpreted manually are basically identical. It proved that the method is effective.关键词：Field image;Local maximum;Local minimum;Directional profile;Rotation transformation2601|337|0更新时间：2024-05-07
- 摘要：The paper introduces a method of semiautomatic extraction of linear objects from aerial images that has good interactive performance, robustness, accuracy and real time response. A linear object is extracted as a consecutive parabola between each input two seed points. First, through fast binary template matching by correlation and optimize algorithm, the initial parameters of the parabola which have eliminated the blunders caused by shadow and other noises are calculated, and then the least squares template matching is used for accurate positioning by evaluating the adjusted precise parameters of the parabola. This paper specifies the mathematical model of least square template matching and the self adaptive template generation to overcome the effects of changes of the width and contrast of the linear feature. The least squares template matching could also be seen as an optimal algorithm that integrates the image feature and geometrical constraints together and has been studied well for long terms. If the proper initial values have been given, the least square method works well and is often applied to solve the accurate positioning problem. The experimental results indicate that the method could gain high accuracy and is robust and ready to practical mapping production from aerial images.2845|315|0更新时间：2024-05-07
- 摘要：Medium reverse figure imaging is still an important problem that puzzles the fields such as earth physics, and especially people focus their eyes on the computer velocity, imaging quality and its stability. First, basic principle of bending rays imaging is discussed briefly in the beginning of the paper, then the reverse figure steps are given in detail, mathematical model for reverse figure is designed too. This paper also presents an improved genetic algorithm to settle the reverse figure of bending rays imaging, and its steps are also displayed out. At the end of the paper, an example that gained by improved genetic algorithm is presented, and in order to pop out its advantages, climb hill algorithm is also compared with it. We analyze the difference of the results from their essences and also point out that the large scale collateral calculation and the restriction of crossbreed and aberrance that lead to the velocity and quality. We can conclude the algorithm improves its velocity, quality and stability efficiently by the curve figures of the improved genetic algorithm and climb hill algorithm.2200|337|0更新时间：2024-05-07
- 摘要：This paper discusses an integral transform that can be applied to region and edge detection. With homogeneity scale and spatial scale parameters, the transform converts image into vector field showing the attractions among pixels in the image, and thus edge detection can be performed by finding divergent vectors in the field. This paper analyses some problem of applying the transform to edge detection and presents a method of estimating homogeneity scale parameter using local information. Based on the transform and previous work, a simplified version of edge detection algorithm is given. Its experimental results are also presented, with comparison to some classical edge detection algorithm. Its edge detection capability is comparable with Canny edge detection algorithm.2608|325|0更新时间：2024-05-07
- 摘要：The Sweep surface is usually described by using the Local moving Frame along the spine. But the construction of the moving frame is strongly limited by the properties of spine. A problem is arisen: can we describe the Sweep surface to avoid using the moving frame? In this paper, a projected based approach is proposed which possesses two advantages: Firstly, the Sweep surface can be presented exactly if spine is line segment or arc curve. Secondly, the Sweep surface can be directly described by projected transform so we can avoid using the local moving frame, then the limitation of continuity of spine for the local moving frame is not necessary. On the other hand, the error control can be simplified to the error estimating of the offset of the spine. Three types of spine, such as line segment, arc curve and NURBS curve, are considered with different methods in the projection algorithm. Furthermore, 3D NURBS offset approximation algorithm is presented for 3D NURBS curve, and the projection algorithm of the sweep surface is introduced. Finally, the robustness and efficiency of the algorithm is verified by a lot of examples.关键词：Sweep surface;Project transform;Moving frame;Approximation;Geometric modeling;CAGD2701|311|0更新时间：2024-05-07
- 摘要：While detecting the property of seed metering device based on computer vision, adjacent objectives are possible to be covered each other, and faster detecting speed is required. In this paper, by analyzing the feature of camera and flat mirror, a calibration method of camera system that is only composed of a single camera and a flat mirror is proposed, and the calibration model of this camera system is expounded in detail. Therefore, the camera system that is only composed of a single camera and a flat mirror can not only raise speed of grabbing image but also spare memory of computer so that the detecting speed is raised. Meanwhile, applying the map relation between image points and special points, an algorithm of reconstructing special point is also given. The calibration method is testified on the experimental table of seed metering device, results show that not only this system can detect overlapping and lost of seeds, but also error is less than 1mm, and the adhesive tape that is used in detection is completely replaced.2214|292|0更新时间：2024-05-07
- 摘要：This paper addresses several key issues for digital earth oriented virtual reality system, with multi layer scene engine (MLSE), view dependent dynamic multiresolution terrain model(VDDMTM), ActiveX component based network model discussed. By means of scene layer, MLSE enables multi variates integration, scene rendering and user interaction. Quad tree structure is used for multi resolution terrain representation. View dependent terrain simplification and multi resolution texture mapping are applied to keep ROI(Region of interest) details and accelerate rendering. Client/Server based ActiveX network model not only meets specific needs for geo data modeling, but also well supports WWW browsers. Based on above research, a prototype system, virtual earth system(VES), was developed to visualize YongDing River and Mountain Himalayan.关键词：Digital earth;virtual reality;Scene engine;Multi resolution;Texture mapping;ActiveX component2605|355|0更新时间：2024-05-07
- 摘要：We know that Line and polygon can be expressed by their feature points because point is the fundamental feature in vector GIS. Because only digital accuracy of point data is defined in the International Geographic Information System(GIS) data quality standard system, so this paper defines the visual indexes of point ellipse, line error band and polygon error donut to assess effect scope of their positional uncertainty according to the probability that the feature points dropping into their error ellipses based on the error ellipse expressing the positional uncertainty in surveying and mapping, we wish to richen and perfect the system of GIS data quality standard system and to control the data quality of many GIS databases to a higher accuracy.2775|375|0更新时间：2024-05-07
- 摘要：In this paper, a novel 1D deformation transformation for handwritten Chinese character shape correction is proposed. From our point of view, handwritten Chinese character can be regarded as a kind of deformable object. Although there are large variations in the same category of character for different handwriting styles, the basic topological structures of them are the same. With a suitable selection of deformation parameters, the 1D deformation transformation could deform a given handwritten Chinese character into 24 different handwriting styles. A deforming parameter controls the deformation degree for each style. The proposed deformation transformation could be applied as a non linear shape normalization method for Chinese character recognition to correct various position deformations such as left slanting, right slanting, top slanting, bottom slanting, too fat or too thin, etc in a handwritten Chinese character. Our experiment shows that the recognition rate can be improved by 1 5% or so by using the deformation transformation, demonstrating the effectiveness of the proposed transformation.3194|468|0更新时间：2024-05-07
- 摘要：In offline handwritten Chinese character recognition, the high variability of the handwriting strokes is the main cause for lowering the recognition performance, thus decreasing the variability of the handwriting strokes is one effective and important way to improve the recognition accuracy. To solve this problem, we propose a new method of optimal sampling features, which are developed from the prevalently used directional features by following procedures. Firstly, four directional factor images are generated from an input binary character image. Next, these four images are transferred through a low pass filter, and then these four low passed images are sampled. The image values at these sampling positions produce a feature vector that is defined as sampling features. In the case of the sampling positions are uniform and fixed, the sampling features are subject to stroke variations, and these stroke variations will increase the within class pattern variability. In order to compensate for stroke variations, the sampling positions should be adaptable to these stroke variations. That is, the sampling positions should be displaced against reference patterns to decrease the within class variability, on the other hand the smoothness of the displacement should be preserved to keep the character's primary structure unchanged. The sampling features satisfying above conditions are defined as optimal sampling features. These two conditions could be expressed as a constrained minimization problem, thus optimal sampling features could be solved in an iteration procedure. For the sake of saving the time cost, a coarse to fine strategy is utilized. Finally, optimal sampling features are obtained, the discrimination of features is increased; and the recognition performance is improved. In order to demonstrate the effectiveness of optimal sampling features, we apply it to the THCHR database and compare it with directional features. The result shows that sampling features achieve higher recognition accuracy than directional features.2948|402|0更新时间：2024-05-07
- 摘要：In this paper, we introduce our research on virtual restoration and gradual changing simulation of Dunhuang frescoes. We present a feasible restoration model as well as introduce the technologies used. The model is designed to give instructions for restoration facsimilers, fresco researchers and fresco protectors. With the assistant of computer, it is a new and more effective way than manual to protect and preserve Dunhuang fresco which is one of the most brilliant culture heritages of the world. The fresco image restoration result can even be used in virtual cave navigation system to show the brilliance of the grottoes in ancient time. Based on the pigment component research of Dunhuang experts, the experience knowledge of Dunhuang artists, and the current well preserved frescoes, we studied color knowledge, image processing,artifical intelligence and etc to restore some color changing, color fading and color stripping Dunhuang frescoes in computer. The most important of the restoration procedure is to separate the fresco image into some recovery objects(In this paper we adopt color layer division and edge detection technology). And then for each recovery object we get appropriate experience knowledge or analogical images for restoration guidance. Of course it still need some image processing tools such as cleaning dirty, color harmonize and etc to improve the last effect of the whole image. With the restored results then we can simulate the procedure that the frescoes change gradually from origin to current status. We made gradually changing demos for each category of restoration. We are satisfied with part of the results, and the achievement is shown in the EXPO'2000 in Hannover, Germany.3165|378|0更新时间：2024-05-07
- 摘要：As the most important technology bass of information industry, digital image proceesing technology has been entering into a brilliant developing era in theory and application under the drive of internet and media application. A varity of algorithms and products have come forth, growing or disappearing out of competition in marketing. Image digital zoom(view interpolation) is one of the most important image procesing technology for digital still camera, which is used to improve the vision resolution and to satisfy demands out of some special processing algorithms. The correlativity theory is the base of image digital zoom(view interpolation)?smoothing and filtering. The spatial and gray correlativity between pixels are studied in the respect of the correlative coefficent, and the spatial distance and gray distance are defined. Based on these results, the digital zoom(view interpolation) and its discrete algorithm are constructed. The emulation result shows that the algorithm is available and efficient in digital still camera zoom function for its least distortion in color and geometry. The discrete algorithm has been used in I.Photo serial digital still camera.2776|352|0更新时间：2024-05-07
- 摘要：On the basic of computer vision principle, a surface measurement system of sheet metal parts is proposed in this paper. Using neural network, the mapping relation between image points and special points is established. Some distorted stripes are obtained on surface, and the points of stripe edges are detected by wavelet edge detection. A searching non supervisor clustering algorithm is discussed, so all of edge points are divided into different groups according to stripe edge situation, the edge points of every group are fitted into a B spline curve. The curves are recognized and marked based on geometric invariance to search corresponding points at sub pixel level. Furthermore, the multi scale and multi resolution attributes of wavelet are applied to image mosaic and data integration, so a large scale surface can be measured. At last the coordinates of points on surface are calculated with the mapping relation between image points and special points, and the measuring precision is less than 0 5mm/m.The system avoid emending optical system distortion of cameras, achieve stereo matching at sub|pixel level, and integrate surface data, so the large surface can be able to measure.2711|475|0更新时间：2024-05-07
- 摘要：In the design of a front end of a real time image fusion system, a high speed, high integrated, general purpose video signal grabbing system has been developed by using one integrated video decoder, a large scale EPLD device. The EPLD device is configured by a program written in the VHDL language. The grabbing system can provide randomly continually data buffer memory access without waiting by using two buffer memory banks, which map to same address space. The grab parameters, for example, the beginning point, the dimension of grab window, can be change on line. It has following advantages: high integration and reliability, small dimensions, simple interfaces and in system programmability, etc. The system can be realized in one EPM7064SLC84 EPLD device without I 2C interface. The full function EPLD chip with I 2C interface is implemented in one EPM7160SLC84 and successfully applied to a multi channel image grabbing and fusion system. The on chip resource usage rate is 80%. On the other hand, the design can be also applied to the other image processing systems as a high integrated video signal grabbing front end without hardware change. The need change can be implemented by change the VHDL program which configures the EPLD device.2726|399|0更新时间：2024-05-07
- 摘要：Industry controlled configuration software has become the mainstream of computer supervisory control system at present and its visualization embodies the quality of software. For further enhancing the visualization of the configuration software, the paper puts forward "online display of the vernier" scheme. The vernier, a rod capable of sliding freely on the curve graph, is designed in a bid to realize data's dynamic online display, thus replacing the monotonous data tables, Lets this special control element to act as the verniers rod that moves about, and make its the length automatically adjustment on the basis of the graph control element of length (width of this special control element has only one or two pixels). This kind of online display method is direct, convenient and easy to be operated with regard to data inquiry and trend analysis. Therefore, the vernier is a reasonable, applicable and original design. "The button overlapping technology" adopted in the course of designing has produced favorable results. The method aims to solve the problem of verniers' display of various length by overlapping a special control element with a graphic one transparently. Besides, the design of size handle and the principle of dynamic data exchanges are also discussed in the paper with suggested methods on their realization.3515|324|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