最新刊期
卷 13 ， 期 9 ， 2008
- 摘要：In this paper we presents a survey on current techniques for real time interactive visualization of massive models. Firstly, the related key technologies are summarized. Then some typical interactive visualization systems for massive models are analyzed and compared. Finally, some possible directions for further research are pointed out.关键词：massive model visualization;real time rendering;visibility culling;parallel rendering;out of core algorithm4812|477|0更新时间：2024-05-08
- 摘要：In this paper, we study a class of degraded images with salt & pepper noise and stick shape, such as crossing. Combining the advantages of wiping off salt & pepper noise by robust method with protecting edge by p(x) harmonic map, we establish a variational model, and then iterative an evolvement equation by the steepest descent to conduct experiments. From simulation results, our method preferably keeps the edge of original image.关键词：robustness;image restoration;p(x) harmonic;degraded image3496|349|0更新时间：2024-05-08
- 摘要：The basis functions of traditional Krawtchouk moments are constructed by the product of two Krawtchouk polynomials of one variable. However, the relation between two directions of the plane is isolated. Hence, new image moments whose basis functions are Krawtchouk polynomials of two variables are proposed in this paper, while a simple method is deduced to compute normalized polynomials. Reconstruction experiments show that, compared with discrete orthogonal moments of one variable, moments of two variables have less reconstruction error.关键词：Krawtchouk moments of two variables;image reconstruction;discrete orthogonal moments3230|249|0更新时间：2024-05-08
- 摘要：FDK algorithm based on single circular orbit can only be used when the cone angle is small. The images reconstructed by Feldkamp algorithm become blurred and distorted when the cone angle increases. As a result there is a limitation for practical application. After the relationship between the Radon transform of the object and the Radon data acquired by the scanning of the single circular orbit is analyzed in the paper and the superposing algorithm is presented based on Grengeat’s formula according to the space variant filtered backprojection(SV_FBP). The first sub result is calculated by applying the FDK algorithm to the projection data acquired by a single circular orbit scan. The second sub result is a correction received by applying SV_FBP algorithm to the missing Radon data acquired by extrapolation. The final result is the sum of the two sub results. The result demonstrates the algorithm is a valid method to reduce artifacts of the images reconstructed by FDK algorithm, and the applicable cone angle range can be 3 to 4 times larger than that of the FDK algorithm. This new superposing method will possess important theoretical and application value in the fields of big and long object reconstruction.关键词：single circular orbit;FDK algorithm;space variant filtered backprojection algorithm;superposing algorithm3480|267|0更新时间：2024-05-08
- 摘要：Image interpolation of cross sections is one of the key steps of medical visualization. It affects directly the results of reconstructed tissues or organs, which plays an important role in medical treatment and diagnose. The traditional interpolation methods are imprecise or of high computational complexity. Aiming at such problems, an interpolation method based on pixel classification is presented. The method classifies pixels of the image to be interpolated by the relativity of corresponding pixels of its neighbor original images. Then the different methods are adopted to interpolate the different points. In addition, error checkout is introduced to check the mismatching points. Experimental results show that not only the complexity of the proposed approach is reduced, but also its quantitative error frequency is less than the conventional methods.关键词：cross section interpolation;medical visualization;classification;matching;Cubic spline3207|231|0更新时间：2024-05-08
- 摘要：Since P M diffusion,Shock filter,coherence enhancing diffusion and their fusing models in existence can not enhance road surface images well, a new model fused by the 3 basic PDE models is proposed. First the relationship between the diffusion strength of coherence enhancing diffusion and the degree of consistency of local directional structure is analyzed, and on the assumption that images are without noise, 3 weight functions depending on the local gradients and the degree of consistency of local directional structure are designed, and the 3 basic PDE models are fused together by the 3 weight functions. Then according to the characteristics of road关键词：road surface images;image enhancement;P M diffusion;coherence enhancing diffusion;Shock filter;fusion3174|271|0更新时间：2024-05-08
- 摘要：Text in video is a very compact and accurate clue for video indexing and summarization. But video texts are usually embedded in complex background, making it very difficultl for text separation from the background information. Hence the OCR accuracy was poor. This paper presents a multi frames based technique to enhance video text image. After extracting a reference text block, we use Hausdorff distance based image matching technique to find and register the corresponding text block. Then the frames average or minimum pixel search method is applied to text blocks to obtain a new text block with a clean background. At last we apply a finite interpolation function to adjust the text block resolution. Experiments conducted on several video sequences show that our enhancement scheme can considerably improve the accuracy of OCR.关键词：video analysis;text tracking;text enhancement;Hausdorff distance3460|360|0更新时间：2024-05-08
- 摘要：It is important to preserve more detailed information in image denoising. In this paper, we extend the total variation model for image denoising is which based on gradient to a new model based on fractional derivative, and then making use of the characteristic that fractional derivative and convolution integral have the same discrete formula in certain situation,We propose a variational model based on convolution integral which can be computed easily. The numerical results show that our model can not only better improve the PSNR of image,but also preserve more detailed information.关键词：image denoising;fractional derivative;convolution integral;texture preserving3823|271|0更新时间：2024-05-08
- 摘要：Contourlet transform(CT) is a method of multiscale geometric analysis, which can result in a flexible multi resolution, local, and directional image expansion. But the Contourlet transform is not shift invariant, that will cause pseudo Gibbs phenomena around singularities in image denoising. In this paper we apply redundant contourlet transform with shift invariant to image denosing, which can capture the intrinsic geometrical structure of image. Meanwhile, we consider the dependencies between the coefficients and their parents in detail. We propose a method of image denoising based on redundant contourlet with bivariate shrinkage rules. The experimental results show that our method can obtain higher PSNR value and better visual effect compared with other methods.关键词：redundant Contourlet transform;image denoising;shift invariant;bivariate shrinkage3512|250|0更新时间：2024-05-08
- 摘要：Speckle noise is serious in SAR (synthetic aperture radar) image. It will greatly affect the information extraction of terra and object and the application of SAR image. A novel filtering algorithm for speckle noise in SAR image is presented here. This algorithm is based on an iterative filter that based on a membrane model Markov random field approximation optimized by a synchronous local iterative method (TSPR). With this algorithm, the affect of the energy function by neighbors’ spatial relation is taken into account. Through self adaptive correcting the penalty coefficient in iteration process better filtering effect can be acquired. According the comparison experiments about faded images caused by speckle noise of various intensities, with the algorithm presented here we can restore the image with higher Signal Noise Ratio (SNR) than TSPR algorithm.关键词：synthetic aperture radar (SAR);speckle noise;self adaptive;filtering3311|284|0更新时间：2024-05-08
- 摘要：A new adaptive motion estimation algorithm based on correlation of motion vectors is presented in this paper. The algorithm utilizes the correlation information of current block’s three neighbouring motion vectors to classify current block into correlation type block and independent type block. Different motion estimation methods are applied for correlation type block and independent type block to achieve lower computational complexity and good estimation accuracy. The simulation results show that the proposed algorithm has much lower computational complexity than other motion estimation algorithms, such as full search (FS), diamond search (DS) and Hexagon based search (HEXBS), with no severe PSNR degradation.关键词：motion estimation;motion vector;correlation;adaptive3145|242|0更新时间：2024-05-08
- 摘要：Robustness and invisibility are used to evaluate the performance of the watermark algorithm. In order to design a complicated scheme with both preferable robustness and invisibility with low complexity, in this paper, a new watermarking based on Integer DCT scheme is proposed. At first we studied the integral transform characteristic of the frequency coefficients,to find out the coefficients which are fit to be embedded and the relation of theirs embedded threshold. Then we formed a new image by extract each block DC coefficient, and select out numbers of the blocks with complex texture. At last, by Integer DCT,the digital watermark signals were embedded into the frequency coefficients of the theses blocks. Experimental results show that the proposed watermarking scheme ensures that the invisibility of watermark, and have good robustness against various attacks such as JPEG compression, noise, filter and so on, even if under the mosaic attack.关键词：digital watermark;integral DCT transformation;visual masking;mosaic attack3344|290|0更新时间：2024-05-08
- 摘要：A new iris recognition algorithm, based on PCA and ICA, is proposed in this paper. Firstly, PCA was applied to the iris images in order to reduce dimension and second order correlation, then ICA was applied to train iris images. In our algorithm, ICA was performed on iris images in the CASIA database under two different architectures, of which one treated the image as random variables and the pixels as outcomes, while the other treated the pixels as random variables and the images as outcomes. The first architecture found spatially independent basis images for the iris. The second architecture used ICA to find a representation in which the coefficients used to code images were statistically independent. No matter which architecture we used to train the iris images, the proposed algorithm was effective.关键词：iris recognition;principal component analysis (PCA);independent component analysis (ICA);unsupervised learning3508|263|0更新时间：2024-05-08
- 摘要：A new preprocessing algorithm for illumination invariant face recognition called Segmented Local Normalization (SLN) is proposed. The main idea is to produce image segmentation so that in each segment, pixel points have similar surface normal distribution and then have similar intensity responses to the light source. Then the local pixel normalization is processed in each segment in order to eliminate illumination. The algorithm firstly establishes Lambert object surface reflection model and secondly a general face surface normal matrix is estimated using SVD. Then the clustering algorithm based on the surface normal directions is used to obtain the image segments, and a local normalization is applied in each image segment. Finally, the traditional face recognition algorithm like PCA is applied on the normalized images. Experimental results based on the Harvard and YaleB face database show that under uneven illumination conditions, the algorithm can increase the face recognition rate efficiently.关键词：illumination;segmented local normalization;face recognition4983|312|0更新时间：2024-05-08
- 摘要：In this paper, a new method for detecting specific objects in the image is presented. It consists of three steps. Firstly, a watershed transformation (WT) is used for over segmenting the image in different small basins and dams. The desired object should be one of these basins or the combination of some ones. Secondly, the basins which are most impossible to be the objects are deleted, and then the neighboring relationships between the remainders are analyzed to construct a region adjacency graph (RAG). Finally, an energy function is constructed and a heuristic genetic algorithm (GA) is proposed to extract the optimized sub graphs from the RAG. The regions corresponding to the sub graphs are the desired objects. Experiment results demonstrate the feasibility of this approach.关键词：object detection;watershed transform;region adjacency graph;energy function optimization;genetic algorithm3508|302|0更新时间：2024-05-08
- 摘要：It is hard to detect lung nodules automatically from CT images for lung CAD system. A detection algorithm is proposed for solitary pulmonary nodules(SPNs) in thoracic CT images in this paper. Firstly, lung field is segmented from original CT image effectively and accurately. Secondly, areas of local maximum gray are found, to segment regions of interest(ROIs) roughly. Then, features of each ROI are extracted, each feature is described quantitatively by the accuracy of SVM classification with each single feature separately, and Mahalanobis distance is weighted by the quantitative parameters. Finally, ROIs are classified to nodule or non nodule with the improved Mahalanobis distance. Experiment results indicated that the algorithm can detect SPNs effectively, it is with relatively high sensitivity and low false neglected rate, and it can provide doctors helpful information to diagnose lesions in early stage of lung cancer.关键词：lung cancer CAD;solitary pulmonary nodule;local gray maximum;weighted Mahalanobis distance;classification3354|237|0更新时间：2024-05-08
- 摘要：Snake model is a kind of deformable image segmentation model based on parameters and has been proved effective to contour detecting as well. It is sensitive to the position of the initial curve, lacks the theoretical guidance to choose parameters and can not deal with the change of topological structure. Snake model is easy to leak out if the edge is weak. This paper presents a modified image force by integrating the region information to improve it. After that, Particle Swarm Optimization (PSO) algorithm is applied to optimize the segmentation results obtained by Snake model. The encouraging results have been shown by experiments with the synthesis images and medical images.关键词：image segmentation;Snake model;image force;particle swarm optimization(PSO)3530|268|0更新时间：2024-05-08
- 摘要：There are two important and arduous problems in 3D human images segmentation. One of them is that the large amount of data volume makes it extremely time consuming. The other is that it is difficult to segment some organs and tissues because the differences of their gray levels are relatively small. This paper proposes an optimization algorithm of 3D segmentation with two thresholds based on improved fuzzy exponential entropy by modifying the maximum fuzzy exponential entropy function, which makes the segmentation much better, and searching the optimal thresholds using the Weighting Immune Genetic Algorithm (WIGA). The experiments on the real thoracic data showed the maximum fuzzy exponential entropy function in this paper obtained better thresholds than the traditional entropy function and the fuzzy function. The searching time of WIGA is about 14 per cent of the complete searching time. Moreover, 100 calculations for thresholds showed the optimization algorithm in this paper was more precise and stable compared with the Simple Genetic Algorithm (SGA) and the Immune Genetic Algorithm (IGA) without increasing the time consumption.关键词：3D image segmentation;fuzzy exponential entropy;immune genetic algorithm3225|236|0更新时间：2024-05-08
- 摘要：Watershed segmentation based on gradient images usually has over segmentation result. To solve over segmentation problem, we propose a new Hierarchical image segmentation method based on Watershed filling and Overlap rate measuring (HWO). Firstly, we transform RGB color space to Lab and statistic the histogram according to a and b dimensions. The watershed segmentation algorithm is applied to 2D histogram and the initial segmentation result is achieved. Then, we associate the segmentation region with the Gaussian distributing, and estimate the parameter value. Finally, we measure the Overlap rate for a hierarchical region merging and get the final result. In the experiment, the two parameters are determined. We then evaluate the segmentation performance with a standard database of human segmented natural images. Results show our method can efficiently solve over segmentation problem, and the combined value of precision and recall measures is 0.609, while is 0.79 when the segmentation is done manually. In addition, the new method also has much less computing complexity.关键词：color image segmentation;2 D histogram;watershed;overlap rate;region merging3301|309|0更新时间：2024-05-08
- 摘要：The standard Mean Shift tracking algorithm is lack of scale adaptation mechanism. A familiar scale adaptation strategy is to determine the scale in current frame by running the standard Mean Shift tracking algorithm three times respectively based on the previous scale, its 10% up and 10% off scales. In this paper, the algorithm was tested on numerous typical scenes and its two drawbacks are found: (1) sometimes it may be stuck in the scale smaller than the real scale; (2) it often responds poorly to rapid scale changes. Such drawbacks can introduce additional scale error and thus increase the risk of missin tracking. Through analyzing above drawbacks in detail, we propose the revised scale adaptation algorithm, in which the criterion of optimal bandwidth selection is modified and adaptive filtering parameter is introduced. Experiment results in numerous scenes show the effectiveness and efficiency of the improved algorithm.关键词：mean shift;object tracking;scale adaptation4080|349|0更新时间：2024-05-08
- 摘要：The current mean shift tracking algorithm is not suitable for the targets that move fast and with entire occlusions and is hard to realize real time tracking when it is applied in the pointing digital signal processors due to much floating point operation. This paper provides the optimal algorithm which contains some improvements of the kernel function, Mean Shift alterative weight and Bhattacharyya coefficient. These improvements not only enhance the capacity of tracking the object which resembles the background but also employ pointing operation to satisfy the need of the real time tracking. Furthermore, the target moving fast and entire occlusions are resolved through combining the optimal algorithm with the Kalman forecast. Now, the algorithm is applied in the TMS320C6416 hardware system and successfully copes with clutter, target occlusions, revolution, scale variations and moving fast in the real time tracking.关键词：mean shift;real time tracking;Kernel function;alternate weight;bhattacharyya coefficient;Kalman forecast4087|301|0更新时间：2024-05-08
- 摘要：Surface reconstruction is one of the classic problems in computer vision. Extracting surface depth from a shaded image (SFS) is one kind of effective method. In this paper, the Lambertian surface model was analyzed in detail and the key factors touched the result was discussed; based on this, an improved algorithm with particular approach and computational method was proved. Experiments show that our algorithm enhance the perception of stereo display.关键词：stereo display;shape from shading;surface reconstruction3201|217|0更新时间：2024-05-08
- 摘要：A novel method is proposed for two step slices registration and fast Shear Warp reconstruction of serial tissue section images. After image pre processing, an automatic stepwise accuracy method was used for slices registration. Based on the elementary result of the principal axes transformation method, the final optimal result was achieved when maximized the mutual information reflected the matching suitability at a given position. Considering the large amount of serial tissue section images, an improved Shear Warp algorithm based on sorted volumetric data structure was applied to reduce the access time of non contribution data cells and consequently speed up the reconstruction process. Experimental results demonstrate that the method is effective and efficient, which is suitable for visualization of serial tissue section images.关键词：image registration;visualization in scientific computing;Shear Warp algorithm;serial section;tissue images3314|216|0更新时间：2024-05-08
- 摘要：Due to the limitation of medical image registration introduced by one kind of feature, a medical image elastic registration method using contour and feature points is proposed. Feature points can be extracted with semi automatic method, which can not only ensure the accuracy of the extraction but also avoid the complicated process of building corresponding relationship of feature points. On the basis of keeping shape of the contour, contour linearization is employed to reduce key points in the extracted contour, which improves the computation efficiency. The sum of the distances between feature points and the distances between contours is chosen as the criterion of image registration. The registration transformation is resolved iteratively by the frame of ICP algorithm. The analysis of the choice of cost function and the experimental comparison with other methods using real images demonstrate that the results of the proposed algorithm are better than those of other methods that only use the point or the contour. The presented method can not only register the structure of images (contour) but also align the interested anatomic positions (feature points), which represents the exact difference between images, and is a fast and accurate medical image registration algorithm.关键词：medical image registration;feature points;contour;iterative closest point(ICP)3888|289|0更新时间：2024-05-08
- 摘要：The traditional co occurrence matrix is an effective approach in texture image analysis. It is widely used in the research sense of image comprehension and computer vision. In order to improve the performance of image retrieval, a new kind of co occurrence matrix descriptor are introduced as novel features for image retrieval. It can describes the spatial correlation of four pixels. During the course of image retrieval using the co occurrence matrix descriptor, color gradient is computed from the RGB color space, then describe the image characteristic by four pixels co occurrence matrix and use to image retrieval is described. Experimental results have shown that four pixels co occurrence matrix descriptor can combine color, texture and shape characteristic, the performances are better than that of gray level co occurrence matrices and color correlograms.关键词：image retrieval;gray level co occurrence matrix;color gradient;four pixels co occurrence matrix(FPCM)3431|256|0更新时间：2024-05-08
- 摘要：A new method is presented for robustly estimating the fundamental matrix from image correspondence. Starting from RANSAC and MLESAC algorithm, we addressed some problems posed from both practical and theoretical standpoints and propose a new algorithm GMSAC. GMSAC adopts the same sampling strategy and maximization likelihood theory as the previous approaches. But instead of uniform distribution used by MLESAC, GMSAC choose Gaussian mixture to model the outliers. Due to the complex nature of outliers’ occurrence, Gaussian mixture is more suitable for the distribution of outliers. We make a detailed analysis on the formation of outliers, and model different types of outliers respectively. Results are given for several image sequences, and it is demonstrated that this method gives results superior to the original MLESAC.关键词：fundamental matrix;outlier;RANSAC;MLESAC;Gaussian mixture model(GMM)3706|267|0更新时间：2024-05-08
- 摘要：A quadrilateral mesh generation scheme using recursive domain bi divisions is enhanced in this paper. Robust mesh templates are designed, to provide a unified way for sub domain meshing. The templates pose no restrictions on the numbers of boundary nodes (e.g. six or eight) of sub domains, and can mesh the ‘big’ regular sub domains bounded linearly in many cases, thus to highly improve the timing and storage performance of the algorithm. Coupling the new sub domain mesh generation procedure with an automatic domain decomposition procedure, we obtain a new automatic unstructured quadrilateral mesh generation algorithm. Finally, mesh examples and simulation results are presented to demonstrate the capabilities of the new algorithm.关键词：mesh generation;unstructured mesh;quadrilateral elements;mesh templates4106|280|0更新时间：2024-05-08
- 摘要：Considering that some helpful scalar properties to vector visualization with feature analysis, in this paper we present an enhanced realtime 2D flow visualization algorithm based on flow lines enhancement. Through the color mapping of some scalar properties, such as velocity, angle and curvature, it can not only reflect flow direction, but also display many properties of vector field which are helpful to analyze the features and topology of vector fields. In order to enhance the contrast among flow lines and improve the rendering image quality, it adopts a scheme of enhancing flow lines which applies one dimensional high pass filter to convolution texture along the direction perpendicular to flow. By using the programmable ability of modern GPVS, the algorithm can achieve realtime rendering performance on Personal Computers.关键词：flow visualization;realtime rendering;scientific visualization;high pass filter3365|253|0更新时间：2024-05-08
- 摘要：Environment registration is a key issue for an augmented reality (AR) system. An efficient markerless registration algorithm is introduced in this paper. The correspondences between frames are obtained in the process of tracking planar target with a fixed template. The homograph of the target can be computed with the perspective projective transformation. Combined with the intrinsic parameters of the camera, pose estimation of the region can be completed resulting in 3D registration for the system. A markerless AR system based on this algorithm is designed. The system architecture and working flow are illustrated. Experimental results demonstrate the accuracy of the theoretical analysis and the robustness of the system.关键词：augmented reality;markerless registration;virtual tracking;motion estimation3670|256|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