最新刊期
卷 15 ， 期 2 ， 2010
- 摘要：Mesh segmentation is one of the important geometry processing problems. In this paper, with attention to geometry signal for segmentation, segmentation type, algorithm principle, algorithm efficiency, we survey those typical mesh segmentation techniques such as clustering method, region growing method, watershed-based method, hierarchical clustering or decomposition method, spectral analysis method, skeleton-based method. Combining with our work, the tendency of this research direction is further discussed.关键词：mesh segmentation;geometry processing;triangular mesh;survey6825|589|0更新时间：2024-05-07
- 摘要：As a non-linear image processing method with two-dimensional convolution operation, mathematical morphology is divided into binary morphology, gray-level morphology and color morphology. Erosion, dilation, opening operation and closing operation are the basis of mathematical morphology. Mathematical morphology can be used for edge detection, image segmentation, noise elimination, feature extraction and other image processing problems. It has been widely used in the field of image processing. Based on current progress, this paper gives a comprehensive expatiation on the mathematical morphology classification and application of image processing of crop disease. In the end, open problems and further research of mathematical morphology are discussed.关键词：mathematical morphology;erosion;dilation;crop disease;image processing4095|443|0更新时间：2024-05-07
- 摘要：The color uniformity is one of the important problems that havent been solved with the color transferring technique yet. Some traditional methods of the color transferring between images are based on the pixels matching. These methods cant ensure that the color of the colorized image feels well. In other words, it means that the color of some pixels is distinct from the color of their neighbors in the colorized target images. This paper presents an image colorization method based on color transfer and propagation. First, it obtains a part-colored image from transferring color to the grayscale image using traditional technique in the lαβ color space. Then it makes the local colors be automatically propagated to produce a fully colorized image. In experiments, the color of the colorized images obtained is very good and the results demonstrate the practicality of this method.关键词：color transfer;lαβ color space;propagation;image colorization4645|477|0更新时间：2024-05-07
- 摘要：In order to restore high quality image rapidly by using multiframe degraded images, a novel recursive projection gradient algorithm for blind restoration of multi-frame images is presented. We take full advantage of priori information of multiframe images. Firstly, we present a new cost function which can efficiently control noise magnification, then minimizes the noise by projection gradient algorithm and calculate iteration formula. Finally we use multiple recursion iterations to reconstruct image in the frequency domain. Results of experiments demonstrate that this method works efficiently. For images of different noise grades and PSF types, it can clearly restore detail features of image and exactly restore degraded PSF at the same time.关键词：projection gradient;recursive;image restoration3894|351|0更新时间：2024-05-07
- 摘要：Geometrical attacks can cause the loss of synchronism between the watermarking embedding and the detection，thus have negative effect on watermark detection. To solve the synchronism problem caused by geometrical attacks, features of the steerable pyramid coefficients are used to estimate the rotation of the watermarked image, then adaptive wavelet tree modulation employing normalized wavelet tree entropy mask and contrast sensitive function is applied to embed the watermark. Owing to the super tree algorithms ability of resistance to scaling attack and adoption of steerable pyramid, the proposed algorithm can resist both scaling attacks and rotating attacks. The simulation result indicated that the proposed algorithm can efficiently resist geometric attacks.关键词：steerable pyramid;wavelet super tree;normalized wavelet tree entropy mask;geometrically robust watermark3761|572|0更新时间：2024-05-07
- 摘要：Digital image scrambling degree is one of the most important aspects of evaluating digital image scrambling algorithm，it has significant effects on the research of digital watermark algorithm. It can describe the best effects of digital image scrambling by taking the frequency of gray levels. Using the best effects of digital image scrambling as parameters, this paper has presented a new method of measuring digital image scrambling degree based on the frequency of gray levels. Emulational experiments have been carried out in MATLAB. The results of the experiment have suggested that the digital image scrambling degree evaluation method is the same as all sorts of image encrypting. The resulting digital image scrambling degree is accordant with visual subjective evaluation, it would not differ a lot for the sake of the changing of its former image as well.关键词：digital image;scrambling transformation;image scrambling degree(ISD);frequency of gray levels3438|305|0更新时间：2024-05-07
- 摘要：In order to improve the quality of HDR (high dynamic range) image, a novel noise removal algorithm is proposed. Firstly, pixel value sequences are obtained from the same spatial position in different frames. Secondly, pixel values are transferred to luminance by cameras ITF (intensity transfer function).Thirdly, the regression beeline L-t (luminance-exposure time) is acquired by solving a least-squared error problem, and then it was used to revise the luminance values of the sequence. Finally, the revised luminance values are anti-transferred to pixel values. Synthesis operation is performed after noise removal. Experiment demonstrates that our algorithm is fast. It can eliminate salt & pepper noise clearly, and can reduce Gaussian noise effectively. It is better than A.&E.s (ahmet & erit). These are the prevailing features of our algorithm.关键词：high dynamic range(HDR);noise removal;least-square error;intensity transfer function(ITF)3872|434|0更新时间：2024-05-07
- 摘要：In many fields such as remote sensing or medicine, lossless compression is usually required because of the demand for high image quality. A lossless image compression algorithm is proposed in this paper considering the limitation of existing methods. The algorithm takes advantage of the effects of the distribution of pixels on compression ratio. The pixels are classified firstly, and the classification results are recorded in mask images. After that, the pixels of each class are re-ordered using the Hilbert curve. Finally the mask image and the data stream of each class are coded using the LZW algorithm. Experiment was conducted on both standard testing images and remote sensing images, and the result showed that the proposed algorithm produced higher compression ratio than the methods of LZW, RLE and Huffman. The proposed algorithm is also easy to implement.关键词：classification;mask image;Hilbert curve;re-ordering;lempel-ziv-welch(LZW);lossless compression3585|439|0更新时间：2024-05-07
- 摘要：Although the search speed of fast motion estimation algorithms such as Three Step Search and Diamond Search is fast, it is easy to trap into local-minimum which caused the false motion vector preditcor. UMHexagons solved the problem well; however, this algorithm is still needed to search too many points and therefore slowed down the search speed. In order to obtain a good balance between search speed and quality performance, this paper proposed PEOSA algorithm.It divided the search area into four part firstly, then evaluated the probability of the best motion vector in every part,and finally searched every part through different search scheme. The experiment results showed that the PEOSA algorithm save more than 43.84% search time compared with UMHexagons; while the averaging PSNR loss was less than 0.017dB for all tested sequences .The new search algorithm holds high value in real time coding and decoding.关键词：H.264;UMHexagons;motion estimation;motion vector3128|334|0更新时间：2024-05-07
- 摘要：Gabor wavelets are one of the important approaches to texture feature extraction. Through the orthogonality of its base functions, the Gabor wavelets can not only extract texture features effectively, but also reduce redundancy. However, the texture feature vector computed from the Gabor wavelets has higher dimension. An approach using modified Gabor wavelets is presented in the paper. The approach uses the Gabor wavelets to compute energy of different scales and different directions, and the dominant peak set. Then the texture feature vector is computed from the dominant peak set. Furthermore, standardized energy is introduced into similarity measure as weights. Experiments show that the system that uses the modified Gabor wavelets has about the same retrieval performance as that uses the Gabor wavelets. However, the dimension of the texture feature vector of the former is only 6.1% of that of the latter.关键词：Gabor wavelets;texture feature extraction;image retrieval6832|575|0更新时间：2024-05-07
- 摘要：In order to effectively segment biological movement image, this paper inspired by the two-dimension maximum entropy, use the clone selection algorithm of computer immunology into the image segmentation. First, this method codes the two-dimension maximum entropy. Second, the fitness function is established according to the criterion function of the two-dimension maximum entropy. Then with the given initial population, we execute selection, cloning, mutation and updating to the population, finally get the best result which can segment image efficiently. The experimental results indicate that this algorithm spend the search time approximately only then the ordinary two-dimensional biggest entropy method 13%. This proof the method is feasible and effective.关键词：image segmentation;two-dimension maximum entropy;clone selection;immune algorithm3794|403|0更新时间：2024-05-07
- 摘要：The columns or rows of an image are practically viewed as sample vectors in two dimension linear discriminent analysis (2DLDA). However, those sample vectors can not fulfill the independent identically distributed requirement in statistics. This paper proposes a method, called Sampling and Regroupment 2DLDA (SR2DLDA), which can improve 2DLDA and LDA. SR2DLDA down-samples the sample images, regroups the small down-sampling images to matrices, and then performs 2DLDA on them. These matrices may make progress on the independent identically distributed requirement. The experiments on ORL database, UMIST database and FERET database verify the efficiency of the SR2DLDA.关键词：2DLDA;Image Sampling and Regroupment;Complete PCA;NLDA3413|346|0更新时间：2024-05-07
- 摘要：A two-step matching method for 3D face recognition is proposed. Feature points are detected based on curvature and geometric constraint. Then the symmetrical plane of 3D face is determined based on the feature points, and the profile is determined by the obtained symmetrical plane. The profile is used to form a rejection classifier, which quickly eliminates a large number of candidate faces at an early stage for an efficient recognition in case of large galleries. The remaining faces are then verified using a novel region-based matching approach, which is robust to facial expressions. This approach automatically segments the rigid regions, which are relatively less sensitive to expressions and it matches them separately using a modified iterative closest point (ICP) algorithm. The results of all the matching engines are fused using a sum rule to achieve higher accuracy. Our simulation experiment on 3D_RMA database demonstrates that the proposed method is simple, efficient and robust.关键词：3D face recognition;profile matching;iterative closest point;rigid region4427|406|0更新时间：2024-05-07
- 摘要：Fingerprint indexing heightens the efficiency of identification to a large extent through reducing the matching space effectively, thereby, it becomes a very important research area in Automatic Fingerprint Identification System(AFIS). This paper mainly includes two work: discovering that the Orientation_based Minutia Descriptor(OMD) is rotation-interrelated and solving the problem by improving the OMD similarity-degree computing method; proposing a new fingerprint indexing method based on improved OMD. The main procedure of our fingerprint indexing method is: calculating the OMD similarity-degree between all the point-pairs which from the current input fingerprint and the fingerprint in database respectively by using the improved OMD similarity-degree computing method, and obtaining the coarse-level correspondence set based on the OMD similarity-degree; constructing the local neighborhood structure to validate the coarse level correspondence further, and obtaining the refined correspondence set; obtaining the rest refined correspondence sets between the current input fingerprint and the rest fingerprints in database by repeating the two fore-mentioned steps; counting the number of correspondences for all refined correspondence set, getting the top N refined correspondence sets and returning their corresponding finger labels. Experiment validated the effectiveness of our method.关键词：automatic fingerprint identification system;fingerprint indexing;fingerprint classification;orientation_based minutia descriptor(OMD)3695|350|0更新时间：2024-05-07
- 摘要：An image analysis framework based on vertebrae segmentation in lumbar MRI is presented in this paper. As PCA lacks for describing data in detail, it is replaced by ICA in both shape and texture modeling with a further fast sorting of Independent Components. Improved ICA-AAMs is then built, with which vertebrae in MRI are segmented. Based on vertebra segmentation, the best fitting function of lumbar curve is obtained with the MDL principle; lumbar lordosis angles are then computed based on the curvature of this lumbar curve. At last, disc degeneration is evaluated from a declining intensity projecting profile between the endplates of vertebrae. Experiments have testified the accuracy and validity of the segmentation and quantitative analysis respectively. This framework reduces the clinical trifle in terms of providing impersonal quantitative information to doctors. It is also extensible to lumbar computer aided diagnosis (CAD).关键词：MRI segmentation;ICA-AAMs;lumbar spine;quantitative analysis3971|360|0更新时间：2024-05-07
- 摘要：Traditionally, bar codes are captured and recognized by industrial scanner In recent years, with the development of mobile value-added services and 3G technology, the application of two-dimensional bar codes in mobile phones develops rapidly Taking Data Matrix as an example, the recognition technology of the two-dimensional bar code for embedded mobile phones is researched in this paper First, a fast bar codes detection algorithm based on chain code tracing and line detection is proposed according to the character of Data Matrix Then the model of blurred bar code signal caused by the point spread function is analyzed, and Wiener filter is used to recover the signal At last, aiming at perspective geometrical distortion, a fast inverse perspective transformation algorithm is designed, which is suitable for embedded mobile phones The experimental results show that the performance of the algorithm proposed in this paper is excellent It can achieve higher recognition rate and faster speed, which suffices to the requirement of the practical use3982|381|0更新时间：2024-05-07
- 摘要：A novel method of implementation for Euclidean distance transform using erosion and dilation with form correction is proposed in this paper. The characteristic of the method is a new data structure, which uses line segment table to represent region and edge. Compared with traditional convolution mask, the efficiency of erosion and dilation operation for region expressed by line segment table is improved more than 10 times. Through analyzing the rules of distortions caused by erosion and dilation, we propose a form correction method to decrease the errors. Compared with the Chamfer algorithm, our new approach is better for fidelity and efficiency. The new distance transform can also be used for image blend thanks to the advantage of fidelity and high speed when creating feather mask. Furthermore, it is especially propitious to feather operation of selected boundary in arbitrarily shape regions. The principle and implementation are discussed in detail. And the efficiencies of different processing methods are compared.关键词：euclidean distance transform (EDT);erosion;dilation;line segment table (LST);line processing (LP);image blend3747|349|0更新时间：2024-05-07
- 摘要：In the video detection system of highway traffic flow, it is difficult to detect vehicles This paper studies nighttime highway traffic vehicles and proposes a robust vehicle detection and tracking algorithm based on optimization theory. The proposed algorithm improves the previous methods for headlight detection and the rules for trajectory tracking. At the same time, it can not only automatically present traffic flow and vehicles speed statistically, but also recognize traffic vehicle event such as jam-packed or driving against the traffic. Experiment results demonstrate the algorithm has lower complexity and better performance than other methods. The detection rate can reach up to 95% or so, robust with low complexity, real-time feature and its detection ratio reaches up to 95% in smooth traffic conditions and 80% in traffic jams.关键词：intelligent traffic;vehicle detection;vehicle tracking;nighttime;highway4205|409|0更新时间：2024-05-07
- 摘要：Particle filter based track-before-detect (PF-TBD) is an active research topic on detection and tracking for dim targets in sequential frame images currently. Whereas there is absent of study on mechanism of PF-TBD in existing work, the principle of TBD which is based on cumulative measurement likelihood ratio to detect dim targets is analyzed in Bayesian inference framework. The course of detection and tracking of dim targets using Bayesian TBD which is based on recursively cumulate measurements is educed and the mechanism which realizes it by unnormalized particles′ weights in particle filter processing is analyzed and confirmed in theory. A PF-TBD algorithm is presented based on this mechanism. Simulation results show that the analysis is valid and PF-TBD algorithm in this paper is efficient.关键词：Bayesian inference;track-before-detect (TBD);particle filter;dim targets3413|476|0更新时间：2024-05-07
- 摘要：An efficient algorithm for scene detection is proposed. Where firstly within a sliding shot window, the shots are clustered and each cluster center is achieved rapidly by employing MS clustering, then according to the development pattern of film video scene, the temporal distances between two shots are computed and scenes are detected based on the temporal and spatial relationship. In addition, the scene key frames can be achieved on the basis of corresponding cluster centers. Finally, a succeeding procession for over-segmented scenes is introduced. Experiments prove this algorithm can cluster shots rapidly and detect scenes and the key frames of the scence efficiently.关键词：mean shift;clustering;shot cluster;scene detection;scene key frame3592|426|0更新时间：2024-05-07
- 摘要：A novel fusion algorithm for medical images using HMT model in aliasing-free contourlet transform (AFCT) domain is presented. First, the frequency aliasing of the original contourlet transform is investigated, and we make sure that the main reason of aliasing is the two lowpass filters of laplacian pyramid (LP) do not satisfy the nyquist-shannon sampling theorem. Then, instead of using LP, a new multiscale decomposition using two channel filter banks by considering a band limiting constraint on the low-pass filter is designed; and combined with directional filter banks, the AFCT is realized. On this basis, a medical image formation model using hidden Markov tree (HMT) to capture the correlations between the coefficients across decomposition scales is proposed. Finally, based on this image formation model, the expectation-maximization (EM) algorithm is used to estimate the model parameters and produce the fused image. The fusion experi〖HJ〗ments have been made on CT/MR and MR-T1/MR-T2, comparing with the traditional fusion methods which is based on wavelet transform and contourlet transform, the proposed algorithm can provide a more satisfactory outcome in terms of visual quality and quantitative criterion.关键词：aliasing-free contourlet transform;filter banks;hidden Markov model;medical image fusion3457|382|0更新时间：2024-05-07
- 摘要：The registration methods based on optical flow can be conveniently implemented and the registration process is efficient. But all of the appeared registration methods based on optical flow use the initial optical flow model, which make the image fuzzed seriously so that it can not be used. In this paper we improve the regularization term in the initial optical flow model and introduce a motion-blurred image restoration algorithm. The new method reduced the blurring of the initial optical flow model. The experiment results show that results of high accuracy and efficiency can be obtained using the image registration method based on improving the optical flow model.关键词：improved optical flow model;medical image registration;regularization term;motion-blurred image restoration algorithm4494|391|0更新时间：2024-05-07
- 摘要：A feature-preserving mesh completion algorithm is proposed. Firstly the octree-based subdivision of mesh vertices is created. Piecewise quadratic polynomial functions are used to fit the mesh vertices near the hole. Around the sharp features, two or more quadratic functions with different coefficients are used so as to capture sharp features such as edges or corners. The extended Marching Cube algorithm extracts the triangle mesh patch which is stitched to the hole boundary of the original model to complete the polygon mesh. Finally a feature enhancement process is applied to eliminate the aliasing artifacts to obtain fine sharp features. The results of experimental tests show that excellent hole repairing mesh can be achieved.关键词：mesh reconstruction;mesh completion;surface fitting3989|385|0更新时间：2024-05-07
- 摘要：By learning the various character image samples, the automatic and synchronistic generation of new Chinese calligraphy styles is a key problem in the computer artistic simulating. A curve analogy method based on FSVM is proposed which can generate new calligraphy styles by users definition. Firstly, the input character image samples are transferred into a hierarchical stroke structural model. Secondly, the matching points in the various skeleton structure are retrieved with FSVM,which provides the control parameter of the curve analogy process. Lastly, the new skeletons are reconstructed into the calligraphy characters. The efficiency of our approach is manifested by the preliminary experiment with the generation of quantity of novel calligraphy styles,which can be manipulated by the user defined setting and input characters styles.关键词：FSVM(fuzzy support vector machines);curve analogy;calligraphy artistic style simulation;precise character skeleton extraction3204|384|0更新时间：2024-05-07
- 摘要：Softness haptic deformation model with high precision and fast speed is a key issue to the realization of haptic rendering system. A novel parallel rhombus-chain-connected haptic deformation model based on physics is proposed in this paper. The accumulation of relative displacements in every chain structure unit is equal to the object surface deformation, and the resultant force of springs is equal to the external force. Haptic display system is established using the hand controller of Delta 6-DOF, then contact deformation of soft objects and real-time virtual haptic feedback is simulated. Experiment results show that the proposed model not only has less calculation but also can ensure synchronously higher precision in calculating contact force and deformation, and is able to satisfy the requirement of virtual reality system for real time performance and accurate operation.关键词：parallel rhombus-chain-connected model;softness haptic display;haptic feedback;virtual reality3544|350|0更新时间：2024-05-07
- 摘要：Segmentation is a basic and pivotal step of object-oriented remote sensing image classification, and the scale is a key problem of image segmentation. Aiming at the optimal segmentation scale selection for object-oriented remote sensing image classification, conformity degree between vectorial boundary lines of image region object after segmentation and true boundary lines of classification objects as criterion, through their multi-dimensions distance to define the conformity degree, the paper brought forward a new method of optimal segmentation scale selection for object-oriented remote sensing image classification-vector distance index method. Research verified the validity and applicability of this method through two experiments One experiment compared the results of optimal segmentation scale selection based on vector distance index method and ‘trial and error’ method. Results showed the vector distance index could reflect the optimal segmentation scale for seven classes exactly. The other experiment classified the segmentation result from the first experiment, through the accuracy assessment, explored the relationship of selection result based on vector distance index method and classification accuracy. Results showed water, dryland, rice, forest and resident region gained the highest accuracy on the scale that was selected by the vector distance index method, although marsh and grass didnt gain the highest accuracy on the scale that was selected by the vector distance index method, through further analysis, the classification result tallied with the practical condition. Then both two experiment results showed that this method could realize the optimal segmentation scale selection for object-oriented remote sensing image classification, which was similar to human thought, unconfined to data source, intuitionistic, comprehensible and practical. Based on the basic theory of vector distance index method, aiming at the ‘submergence’ and ‘fragmentation’ phenomenon, research brought forward a scale index, which could reflect the small or big status for a given object type, and provided a quantitative tool to assess the conflict degree between ‘submergence’ and ‘fragmentation’, then showed its significance during the process of segmentation scale selection.关键词：object-oriented classification;remote sensing image;segmentation;scale;vector distance index6518|423|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