最新刊期
卷 16 ， 期 1 ， 2011
- 摘要：A fast rate allocation algorithm of joint source and channel coding is proposed for progressive image transmission on band-limited and time varying channels. Firstly the channel is segmented into several operating regions according to the error-probability performance of a set of rate-compatible channel codes. In each operating region, the number of available channel codes in the rate set is reduced, and the optimal error protection scheme is obtained by forward and backward rate searching. The proposed algorithm has a low complexity, and its computation number is only ten percent of that of the heuristic search algorithm, thus the operation time is reduced. When the fast algorithm is applied in adaptive transmission systems, the rate allocation can be quickly adjusted according to channel conditions. Simulation results show that the PSNR value of reconstructed image is always above 29.5 dB and the fluctuation range is less than 4 dB, which shows a good and stable transmission quality.关键词：progressive image;joint source and channel coding;adaptive transmission;fast rate allocation algorithm3638|256|0更新时间：2024-05-08
- 摘要：In this paper, a direction-based adaptive mix-pattern motion estimation search algorithm is proposed. The algorithm makes the best of spatial and temporal correlation, which improves the veracity of the search center prediction compared with the algorithms employing spatial correlation only. By employing adaptive threshold technology on static block determination, the disadvantage of const threshold technology on static block determination has been avoided. Moreover, the algorithm chooses the search strategy according to the corresponding motion type, which can balance the video quality and speed simultaneously. Experimental results show that the proposed algorithm can highly reduce the time of motion estimation ensuring the video quality compared with the UMHexagonS algorithm.关键词：H.264;motion estimation;adaptation;block matching3576|277|0更新时间：2024-05-08
- 摘要：An algorithm for medical X-ray image enhancement based on human visual properties and improved PCNN image factorization is proposed. Using the improved image factorization algorithm, an image is decomposed into a set of image factors which are ordered from coarse to fine in details. Each factor is separately smoothed to obtain a gain matrix and then the matrix is adjusted adaptively according to whether its local contrast reaches the contrast threshold resulting from human visual properties. The enhanced image is reconstructed from the enhanced factors. Through simulations to different medical X-ray images, a better effect is achieved compared with common image enhancement methods.关键词：pulse coupled neural networks;image factorization;medical image enhancement;human visual properties3232|283|0更新时间：2024-05-08
- 摘要：To overcome the shortcomings of temporal error concealment techniques in decoder that can not estimate motion vector of the loss block accurately, motion estimation in decoder and implicit motion model combined error concealment scheme is presented. First of all, we utilize the survived surrounding pixels of lost block for motion estimation, and determining the accuracy of estimates. Then we restore the missing pixels with implicit motion model making full use of the relevance of temporal and spatial domain according to the accuracy. The simulation results show that this method can improve the video image quality by 1dB than other methods.关键词：video encoding;error concealment;motion estimation;virtual camera;implicit motion model3156|252|0更新时间：2024-05-08
- 摘要：Contour scanning based Euclidean distance transformation (EDT) is a fast and complete EDT algorithm and achieves the optimum complexity. But it needs to search all contour regions, so the computation is not optimal. To overcome this disadvantage, an EDT algorithm based on contour hierarchical scanning was proposed. First, reclassifying pixels of the binary image according contour character of central pixel is needed. Second, according the spatial relationship between contour region pixel and central pixel, searching contour region of central pixel hierarchically is the main step to reduce the computation. Finally, we use the proposed terminal condition ends the searching. Theoretically, the proposed algorithm maintains the optimum complexity and reduces the computation. Likewise, the experiments showed that the proposed algorithm can obtain the Euclidean distance images exactly and reduce the calculation time cost.关键词：hierarchical searching;contour scanning;complete EDT3380|253|0更新时间：2024-05-08
- 摘要：A robust background subtraction technique is proposed based on adaptive clustering of temporal color/intensity. An un-supervised clustering method is proposed to model a background with a group of weighted clusters. The clusters and their weights can be updated with a background change. In addition, the unimodal or multimodal distributions of background are detected adaptively. We also present a novel statistical threshold estimation scheme to determine the thresholds using in our method. Experimental results on different types of videos demonstrate the utility and performance of the proposed approach.关键词：background subtraction;kernel density estimation;statistical threshold estimation4499|399|0更新时间：2024-05-08
- 摘要：Accurate extraction of the vasculature in medical images is prerequisite to structural analysis and further applications such as surgical planning. Region growing algorithm is a simple and effective method to extract thick blood vessels which makes use of the spatial continuity of the vascular tree, while the extraction result of small vessels like hepatic artery is unacceptable. In order to solve the problem that the continuity of tenuous vasculature is poor in medical images and vessel segmentation based on traditional region growing may lose distal branches, a directional region growing (DRG) algorithm is proposed which can skip the low gray area in the vasculature during the growing process. The algorithm grows towards the direction of the maximum gray around the grown region, and adds one voxel to the grown region in each iteration. The image is transformed into a tree after the growing process in which the seed point is the root. A trace back procedure beginning from the leaf nodes of the tree can finally determine the region of interest (ROI). The algorithm relaxes the conditions to determine ROI, and small area with low gray in the ROI is permitted. There are two time-consuming steps in the algorithm due to the enormous amount of data in 3D medical images, one is to determine the growing direction in each iteration, the other is to construct the paths from the seed point to leaf nodes during the trace back procedure. Data structure to improve the speed of the algorithm is discussed. The algorithm can be applied to images with any dimension. The algorithm is tested with 2D and 3D images. In both conditions, the segmentation results obtained by DRG contain more distal branches in comparison with traditional region growing algorithm. To some vein phase CT images with poor quality, the proposed algorithm can also generate better results. Four parameters should be appointed in the algorithm and the empirical values are given. The computational time of the algorithm on 3D images is several seconds, which is acceptable in clinical applications. The surface of the extracted vasculature is rough due to the discrete nature of digital images, and further study is needed to smooth the surface before visualization.关键词：image segmentation;region growing;vessel segmentation4886|333|0更新时间：2024-05-08
- 摘要：Classic generalized Hough transform (GHT) can locate non-deformed shape object, while it is difficult to solve the problem when the target is similar but not necessarily identical to the user or deformation. A two level deformed target locate algorithm based on variant of the well-known GHT for solving this problem is presented. Firstly a two-level locationscheme from coarse to fine strategy is introduced to reduce search range from whole image space, in coarse location step, the edge local binary pattern (LBP) histogram features are extracted to detect the range of the target. In fine location step, making use of the edge points of the image detected and the R-table obtained from the template image to search the feasible parameter, and a large dispersion window used to merge vote results because there is non-perfectly aligned points near the optimal parameters. The experiment results demonstrate that the method is effective to the deformed target locating while time and memory cost is much less, and it is valuable in many applications.关键词：generalized Hough transform;local binary pattern histogram;object location;accumulator matrix3403|239|0更新时间：2024-05-08
- 摘要：In the last few years linear Discriminant Analysis(LDA)become more popular.But small sample size problem(SSSP) is always the biggest problem to perform it.In order to overcome this shortcoming,2D-LDA is proposed. We improve 2D-PLDA by Pre-Clustering,which can make the distribution matrix precisely;then a new iris recognition arithmetic is proposed which is combined with 2D-PLDA and wavelet transform.In the experiment,preprocessing was performed at first,then we extract the feature vector from the known class sample by 2D-PLDA and wavelet.In the validation step,we use Euclidean distance and Hamming distance to find the K-nearest neighbor to decide which class the unknown sample belongs to.From the experiment result,we can conclude that,the proposed arithmetic can achieve higher recognition rate than traditional LDA and 2D-LDA, also the new arithmetic is simple in calculation. We use the different subbands as the input space of 2D-PLDA,so the robustness is enhanced.关键词：iris recognition;2D-LDA;2D-PLDA;wavelet subband;feature matrix2981|276|0更新时间：2024-05-08
- 摘要：Feature extraction is an important and critical step in the process of face recognition, a feature extraction algorithm based on uncorrelated discriminant locality preserving is proposed, and its application in face recognition is carried out. Laplacianface which is based on locality preserving projection is an effective algorithm, but it only takes the locality into account, thus the extracted features might be highly correlated, and it does not consider the class information. Some improved algorithms consider the discrimiant information, but the interclass information is not considered. The algorithm proposed here not only imposes an uncorrelated constraint to reduce data redundancy, but also utilizes the class information and the interclass separability after projection is enhanced. Experiments validate the correctness and effectiveness of the algorithm, and prove that the proposed algorithm has better performance.关键词：feature extraction;locality preserving projection;uncorrelated-ness;face recognition3052|277|0更新时间：2024-05-08
- 摘要：Local characteristic scale of images plays an important role in feature extraction and local scale invariant construction. This paper proposed a local characteristic scale selection method based on Gabor wavelets, in which a Gabor scale-space kernel was formed using the 2D Gabor function model of simple cortical cells. A Gabor scale-space representation of the image was calculated based on the kernel first, then the maximum over scales of the feature point was detected and the scale corresponding to the maximum was selected as the characteristic scale. Experiments results show that the proposed method can select the characteristic scale effectively for different types of features in different conditions. Comparison with LoG method shows that the proposed method has better accuracy, applicability and reliability.关键词：local scale selection;Gabor scale-space;Kernel function;simple cortical cells;LoG (Laplacian of Gaussian)4400|317|0更新时间：2024-05-08
- 摘要：This paper proposes a rotation-invariant algorithm based on NSCT against the common rotation problems for image retrieval. By calculating the mean and standard deviation of image subbands decomposed by NSCT, the texture feature elements are extracted. For each scale, the feature elements are re-ordered ascending by the sum of mean and standard deviation of each image subband, and the orientation sequences are adjusted by considering rotation-invariant to create orientation-weights and feature element-weights. Weighted euclidean distance was used to improve retrieval efficiency. Experiments were conducted on the image set produced by Brodatz database, and the results demonstrate superiority of the proposed method.关键词：image retrieval;non-subsampled Contourlet transform (NSCT);re-order;rotation invariant;weight3767|262|0更新时间：2024-05-08
- 摘要：Outliers are inevitable in image matching process. To address this issue, a novel topology constraint based outlier rejection algorithm is proposed to efficiently remove the mismatches between images after coarse matching. By using the topology constraint to filter the sample sets, the proposed algorithm calculates the transformation between images based on the sample set which fully satisfies the topology constraints. Experimental results demonstrate that the proposed algorithm can significantly reduce the computational complexity, while keeping the accuracy compared to the traditional RANSAC and improved PROSAC algorithms. Therefore, the proposed method can effectively and efficiently improve the performance of image matching, and furthermore benefits the application of 3D scene reconstruction in both accuracy and robustness.关键词：image matching;outliers;order constraint;affine invariant constraint;topology constraint4304|356|0更新时间：2024-05-08
- 摘要：In this paper, a new bilateral registration measurement based on regional information and local frequency information is presented. Firstly, the original image is considered as the reference image and the forward regional mutual information as well as local frequency information is calculated. Meanwhile, the input image is regarded as the reference image and the corresponding backward information is obtained. Then, the forward and backward information is summed up and the residual information is subtracted from the summation to get the final bilateral similarity measurement that can be used for registration. Experimental results on real medical images indicate that the measurement is not sensitive to noise and low sampling resolution and can effectively improve medical image registration precision.关键词：bilateral medical image registration;regional mutual information;Gabor filter;local frequency information3113|277|0更新时间：2024-05-08
- 摘要：This research aims to assist doctors to detect micro-calcifications. In connection with the similar lesions retrieval problem of micro-calcification cluster in mammography, we develop a new algorithm with multi-feature fusion and relevance feedback based on the study of single feature and feature fusion using single distance measure image retrieving techniques, this method adopts multi-distance measure to calculate the similarity directing at different features. Experiment is based on mammography image database which contain 250 mammography images and each image contains calcification cluster, we verified the retrieval performance by the precision-recall ratio (PVR) of single feature, feature fusion and relevance feedback. Experimental results show that the method has a better retrieval result than these methods which are single feature and feature fusion based using single distance measurement.关键词：mammography image;calcification lesions;content-based image retrieval;feature fusion;relevance feedback;multi-distance measure3377|257|0更新时间：2024-05-08
- 摘要：Aiming at the problem of low efficiency and mismatching caused by occlusion in global stereo matching, this paper proposes a novel method of stereo matching with high efficiency, which is based on hierarchical belief-propagation. First, this method obtains initial matching cost from local stereo matching, and occlusion and mismatched pixels are checked out by initial disparity. Then, the result of plane estimation is used to correct the initial cost, thus pixels in occlusion region can be refined. At last, in order to reduce the mismatching in low texture region and enhance the smoothness, an improved hierarchical belief-propagation is used to estimate the minimal global energy. Experimental results demonstrate the outstanding performance of the proposed method in accuracy and matching speed.关键词：global stereo matching;disparity estimation;hierarchical belief-propagation3775|260|0更新时间：2024-05-08
- 摘要：This paper proposed a method of generating video mosaic panorama in single autonomous PTZ tracking application, and this method can compose panorama video of an target object moving on large background image. Our approach can be potentially used in single PTZ surveillance and helpful in monitoring moving object in large scenes. The video mosaic panorama can be separated into target object and panorama background. We first utilize mean shift tracking algorithm to obtain the target region and then align adjacent images by combining vertical projection and harris corners of images, so that image registration can be acquired in real time, meanwhile the transforming parameters to background image are recorded. In the end the whole video mosaic is retrieved by superimposing the target region on the panorama background frame by frame. Compared with traditional method, this can be accomplished without manipulating camera or aligning frames by hand.关键词：autonomous PTZ tracking;panoramic video;panoramic background;visual tracking3593|318|0更新时间：2024-05-08
- 摘要：In order to develop the theory of B-spline curve, the representation of cubic uniform B-spline curve is extended to a general form based on linear combination of curves. Moreover, some properties of the extended curve are discussed in details. The research shows that the basis of the generalized curve is relative simple, and the extended curve includes the original B-spline curve and shows much better shape-control capability than the original curve. Meanwhile, the extended curve keeps the same degree of original one. It is easy to find that the curve can be reshape by adjusting the shape factor. Also, a new method of the representation of closed curve is given.关键词：curves;linear combination;B-spline;shape factor;extension3757|258|0更新时间：2024-05-08
- 摘要：The key in clipping algorithm is the efficiency which is mainly influenced by computing the intersection points between the clipping window and the clipped object. Particularly, for line clipping with respect to elliptical window, to compute the intersection points between the ellipse and the line the quadratic equation has to be solved which involves the extraction of square root that is considered inefficient. To address this, we develop a technique of 5-bit region encoding by which the relationship between an ellipse and a line segment can be determined quickly and accurately. The line segments that are completely visible or completely invisible are discarded; the line segments that surely intersect with the ellipse are dealt with by middle-point segmentation algorithm to obtain the approximate intersection points; and the remaining line segments are clipped by solving the quadratic equations. The proposed algorithm is much more efficient than traditional algorithms, and is straightforward.关键词：line clipping;ellipse window;region encoding;middle point3168|212|0更新时间：2024-05-08
- 摘要：Non-negative matrix factorization (NMF) method is an effective method which decomposes the nonnegative matrix into two non-negative factor matrices. New iterative formulas of the two non-negative factor matrices are proposed based on the exponent distribution in this paper. The bases acquired by this method are disordered. But the order is very important for target detection. A new NMF approach combined Mutual Information is proposed, which used for the detection of SAR images. In the algorithm, priori knowledge of targets is used to obtain the favorable feature vector groups,the feature maps about the test image are respectively gained by using the feature vector groups. All the feature the maps are weighted into a general characteristic map. Finally, the targets are extracted in the characteristic of the map by choosing a suitable threshold. Experimental results of ADTS target high-resolution airborne SAR data show that this method is effective and feasible.关键词：non-negative matrix factorization;mutual information;feature selection;target detection;synthetic aperture radar image3182|249|0更新时间：2024-05-08
- 摘要：A modified information cut algorithm(MIC) is presented. First, information cut(IC) model is demonstrated to be equivalent to Cauchy-Schwarz cut(CScut), and then the optimal solution of IC objective function using graph spectral method is proposed; Using both the gray and space relationship of pixels in an image, a MIC algorithm is proposed based on IC algorithm, this method firstly utilizes Parzen windowing function that combines gray information and space information to evaluate probability density functions, and thus reduces the effect of gray changes to image segmentation. Experiments using synthetic image with noise and remote sensing images indicate that MIC algorithm has better anti-noise performance than IC algorithm, and lower computational complexity compared with graph spectral methods.关键词：image processing;information cut;graph spectral method;Parzen windowing;remote sensing image segmentation4116|263|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