最新刊期
卷 11 ， 期 9 ， 2006
- 摘要：Wavelet domain image denoising has been a striking research problem in image processing. In order to give people a summary knowledge of wavelet domain image denoising, based on many literatures of wavelet domain image denoising, this paper attempts to make a survey of wavelet domain image denoising. First, it gives the characteristics of wavelet transformation, and gives the optimization criterion of wavelet domain image denoising, the selection of wavelet basis. Then, describes the methods of image denoising, and analyzes the building model methods capitalized on wavelet coefficients. At the end, the future trend of wavelet domain image denoising is pointed out.关键词：wavelet domain image denoising;Bayes principle;priori probability;wavelet coefficient model3242|491|0更新时间：2024-05-07
- 摘要：H. 264/AVC video coding standard employs multiple coding-modes and significantly improves coding efficieney at the cost of heavy computational load. To reduce the computation burden, a fast mode decision algorithm with adjustable computational-complexity is presented. According to the analysis of the characteristics of the coding-modes and statistical characteristics about the best mode, the proposed algorithm exploits motion activity and the quantized transform coefficients of the residue obtained by the tested coding-modes, to effectively predict the coding efficiency of the remaining modes and skip the unnecessary modes, further more thresholds used in this algorithm can be adjusted to regulate computational complexity. Experimental results show that the speed of the proposed method is about 2 -4 times faster than the full-mode decision algorithm without any noticeable loss of image quality and coding efficiency.关键词：video coding;H. 264/AVC;mode decision;coding efficiency2711|285|0更新时间：2024-05-07
- 摘要：A real-time single-pass bit rate control algorithm in low bit-rate of macro-block layer for H.264 is presented.In order to solve the drawback of rate control model in H.264,a new concept of average pixel information bit(APIB) and the novel rate control model based on it are represented at first;then the more precise bit allocation scheme is implemented according to the coding complexity of each macro-block,and the parameters of rate control model are updated by the change of APIB adaptively;finally,the complete rate control algorithm is proposed.This algorithm is implemented in JM85(H.264/AVC),the experimental results show that,compared to JVT-G012 in H.264,the proposed method gains 0.171kbps in control precision and 0.227dB in PSNR averagely,and controls the fluctuation on bit and PSNR precisely at the same time.关键词：bit rate control;H.264 video coding standard;average pixel information bit(APIB);macro-block2572|220|0更新时间：2024-05-07
- 摘要：The third generation mobile system(3G) makes mobile conversational video transmission in wireless environment possible and H.264/AVC is optionally supported by the latest 3GPP/3GPP2 standards.Error resilience protection for 3G video stream is necessary due to high packet loss rate in wireless environment.Moreover,3G mobile terminals support only part of error resilience tools of H.264/AVC because of the hardware restrictions.This paper presents various error resilience tools and their usability in 3G environment,at the same time,some simple error resilience schemes for packet-switched 3G conversational services are proposed.Performances of these strategies are tested using off-line common test conditions.Experimental show that these strategies can obtain reasonably satisfactory error correction ability.Experiments data also show that,encoder with low complexity FMO method and appropriate intra block refreshing can get better error correction ability,because this method can make full use of spatial correlation between frames and reduce the requirement of bit rate.2521|196|0更新时间：2024-05-07
- 摘要：Existing work in outlier detection emphasizes the deviation of non-spatial attribute not only in outlier detecting in statistical database but also in spatial outlier detecting in spatial database.However,both spatial and non-spatial attributes must be synthetically considered in many applications,such as image processing,position-based service.We defined outlier in respect of taking account of both spatial and non-spatial attributes and proposed a new density-based spatial outlier detecting approach with leapingly sampling(DBSODLS).Existing density-based outlier detection approaches must calculate neighborhoods of every object,which are time-consuming.This method makes the best of neighbor information that have been detected,leapingly selects the next object, but not every object,which reduces many neighborhood queries.Theoretical comparison shows this method is better than other density-based methods in efficiency,and the experimental results also show that the approach outperforms the existing density-based methods in efficiency.关键词：data mining;spatial outliers;spatial database;impact neighborhood2454|208|0更新时间：2024-05-07
- 摘要：Dual-energy X-ray imaging system combining the high and low transmission signal to serve the method of object classification(calculate R related with effective atomic number) at present.But(object's) thickness has great effect on dual-energy transmission imaging.Videlicet,(object's) thickness influences the veracity of object classification.This paper put forward an improving numerical optimized method by the ratio of two transmission signal to eliminate the effect of the thickness,with optimized area absorption coefficient transmission signal to replace the transmission signal.This method can weaken the effect of(object's) thickness on R,decrease the probability of inaccuracy and improve the veracity of classification by analyzing and evaluating experimental data.关键词：dual-energy;X-ray detect;security inspection;effective atomic number;object classification;energy selection2490|195|0更新时间：2024-05-07
- 摘要：Linear discriminant analysis based on Fisher criterion is one of effective methods for feature extraction, and it was successfully utilized for face recognition. But face image data distribution in practice is highly complex because of illumination, facial expression and pose variations. So it is necessary to extract nonlinear features for face recognition. A novel method called subspace discriminant analysis based on kernel trick is presented in this paper. In the new approach, the kernel trick is used firstly to project the original samples into an implicit space called feature space by nonlinear kernel mapping, then two equivalent models based on generalized Fisher criterion have established by the Theory of Reproducing Kernel in the feature space, and the optimal discriminant vectors are solved finally by using the technique of orthogonal complementary space. The proposed algorithm was tested and evaluated on the ORL face database and the NUST603 face database, which can reach recognition rate shuch as 94% and 99.58% , respectively. The experimental results show that the novel method outperforms both KPCA in[8,9] and Kernel fisherfaces in[ 13] and is comparable with the method in [ 14 ] in terms of correct recognition rate.关键词：Fisher linear discriminant analysis;Kernel function;orthogonal complementary space;human face recognition3052|202|0更新时间：2024-05-07
- 摘要：Human face tracking has received extensive attention because of its potential applications in many fields,such as video coding,surveillance and human-computer interface.The aim of this paper is to track human face in color image sequences effectively.A method using human skin feature integrated with Condensation(conditional density propagation) algorithm is proposed.At first,this paper presents a method of self-skin segmentation based on Hsu R L's skin color model,to detect human skin in color image correctly and effectively.Compared with conventional skin color model methods,this method has more precise detection performance in varying lighting condition.Additionally,this paper presents two improvements for Condensation algorithm: a simple and efficient sampling method based on the Metropolis algorithm and an adaptive dynamical model are used in Condensation algorithm for robust face tracking.Experimental results show that in complex backgrounds and occlusion case,this approach could track human face's free movement.关键词：color image sequence;Face tracking;Condensation algorithm;self-skin segmentation2377|213|0更新时间：2024-05-07
- 摘要：The traditional correlation matching method has poor performance with large biases in the presence of the(object's) large scale change and rotation angles.So,in order to get more high precisely locating result,proposed a robust object locating method based on log-polar transform and affine transform which can locate the object misaligned by scale and rotation changes to the template.Experimental results showed that the algorithm not only causes high precisely locating result,but also more robust relative to the traditional correlation matching method.关键词：log-polar transform;affine transform;correlation matching3445|217|0更新时间：2024-05-07
- 摘要：Recognizing people by their gait is a recent research hotspot. Compared with other biometrics,gait has the following three advantages: distance recognition,uninvasive and difficult to conceal.A lot of research on the gait feature extraction has been done,but some of them are much expensive computation,while others are poor identification effect.A new fast gait recognition method based on the region object area is proposed in this paper.Firstly,we divide the two dimensional silhouette of the walker into three regions(head region,trunk region and legs region).Then their region object areas are computed respectively.Together with the ratio of the (silhouette's) height to width,the gait feature vectors are constructed to identify different subjects.Moreover,a kind of N-best classifier is improved in this paper.Experimental results show that the new method is not only simple and fast,but also very effective.It can also be used on different database.Recognition rate of approximately 90% on the UCSD database and over 98% on the CMU database are achieved.关键词：gait recognition;region area;N-best classifier2574|186|0更新时间：2024-05-07
- 摘要：Discrete grid wavelet transform has wide application in the engineering analysis. The way to decide the value of the scale factor widely used in present is to set the scale fact or aasa=an0(a0=2, n=0,1,…,N). But for the actual applications, it is too rough for analysis. Therefore, people should reduce the scale factor interval to increase the density of the discrete grids of the wavelet transform. However existing ways to increase the density of the discrete grids have some what of randomicity. Through the research in the wavelet transform profilometry, Weng Jiawen and Zhong Jingang proposed another new approach to the analysis of the values of the scale factor in the light of the energy characteristic of the wavelets. In this paper, they took the Gabor wavelet for example and figured out the length of the frequency border of the Heisenberg box for each“daughter wavelet”. Therefore they got the theory formula of the decision of the scale factor for the Gabor wavelet asa=1.15 na0. The theory and the accuracy check by a simulation and an example of 32D shape measurement are explained and shown in this paper. And the results of the simulation and experiments have shown the advantage of the presented method.关键词：wavelet transform;discrete grid;scale;Gabor analytic wavelet3148|213|0更新时间：2024-05-07
- 摘要：This paper presents a novel multi-resolution based approach which utilizes both the color and the edge information to create mosaic from two neighbor images. Unlike current image stitching algorithm, which usually requires pure horizontal camera panning, more overlapping and less exposure differences between two images, our method avoids the strict constraints on how the images are taken and the quality of the images. The proposed algorithm is robust because it is less influenced by the illumination differences and the distortion of the images brought in by lens. Experiments indicate that our algorithm is fast for the introduction of multi-resolution method.关键词：panorama;image mosaic;image registration;multi-resolution3409|242|0更新时间：2024-05-07
- 摘要：The imaging system of SAR is totally different from optic's,thereby,the traditional matching method can not be fully suitable in the case of SAR image process.Based on the development of automatic registration of SAR images,in this paper,a novel method of automatic matching of tie-points with high-resolution SAR is proposed.The method consists of three steps: In the first step,back-trace is employed to determine the type and parameters of the initial mapping function after pyramid images are created;secondly,layered back-trace is used to increase control points,the type and parameters of the optimal mapping function are resolved afterwards;finally,the coordinates of control points are revised on original image resolution in order to acquire the final matched tie-points.The strategy of layered back-trace is employed in this algorithm, which not only resolves the problem of compute efficiency,achieves a higher matching accuracy with more evenly distributed tie-points, but also insures the mapping function to be optimal in MMSE(mean minimum square error) sense.High-resolution airborne SAR images are used in our experiments, and the result shows the proposed algorithm is efficient and effective,where comparing with a traditional method which costs 221s computer time,our method only costs 34s;what is more,it can achieve a sub-pixel matching precision,equaling to 0.284 636 pixels.关键词：layered back-trace;matching of tie-points;high-resolution SAR3630|217|0更新时间：2024-05-07
- 摘要：In order to reconstruct accurately for multi self-calibration images with marked point occlusions,this paper proposes a method for reconstruction and camera poses recovery from multi self-calibration images with marked point occlusions.Compared to previous methods,this method uses a way of marked point: coded point and non-coded point.The coded points are utilized for a single CCD camera self-calibration and camera poses recovery and the non-coded points are utilized for 3D points reconstruction.The novelties of the proposed method are three-fold.Firstly,the problem of manually selected image point correspondences is avoided by using auto match of marked points.Secondly,the proposed method is based on marked points,the precision of 3D reconstruction is improved since match of marked point is accurate.Thirdly,the robustness is markedly increased since the effect of noise on the marked points of images is small.The practice shows that the results are accurate and reliable so as to meet the requirements of reverse engineering applications.关键词：structure from motion;marked point;camera self-calibration;missing data2512|225|0更新时间：2024-05-07
- 摘要：In this paper, it is discovered that the state of Martin process on sine function varies with the parameters. The system state varies from steady focus and multiple bifurcation to strange attractors, moreover this process repeats once again. The system is further shown to be chaos quantificationally by computing the largest Lyapunov exponent and drawing phase portrait of this system.关键词：Martin process;attractor;Lyapunov exponent;chaos2449|169|0更新时间：2024-05-07
- 摘要：Curve reconstruction from a set of unorganized points plays an important role in the fields of reverse engineering and computer vision.In this paper,an adaptive genetic algorithm is presented for curve reconstruction based on the regular distribution property of the point set.In this paper,the algorithm divides the region of point set into many grids and in each grid the algorithm elects a point by the adaptive genetic algorithm.Because of the asymmetry of the date points in each grid,here the sphere-of-influence graph(SIG) with several extensions was investigated,which provides a natural notion of proximity in our context,to adjust each data point which have been elected.This paper make these adjusted data points in order,then can reconstruct the curve using B-spline curve based on these ordered data points.The algorithm presented is feasible and simple,specially not only in two-dimensional situations but also in three-dimensional situations.As to the self-intersect point set and the point set having corner point it can also work out a satisfied result.关键词：unorganized points;curve reconstruction;adaptive genetic algorithm;sphere-of-influence graph(SIG)3378|205|0更新时间：2024-05-07
- 摘要：Due to the effect of a sensor,atmosphere etc,remote sensing images are degenerated.To improve the quality of remote sensing images, this paper proposes an image restoration approach based on the modulation transfer function(MTF) according to the analysis of the image degeneration theory.Our method mainly includes two parts: denoising and MTF compensation.An effective denoising method on the frequency space is presented according to the characteristics of remote sensing images.An exponent adjusting method of MTF curve is presented to adjust the MTF compensation degree.The experimental results with CBERS images proved that our method is effective to improve the image factors,such as image contrast,entropy etc,so as to compensate degenerated images.关键词：image restoration;modulation transfer function(MTF);denoising;CBERS images3316|241|0更新时间：2024-05-07
- 摘要：A new improved thinning algorithm is proposed in the paper. In the new algorithm, we advance a series of combination-templates which reinforce the speed of templates matching, and then optimize the combination templates. Finally based on the optimized templates, we bring forward the new algorithm. After doing lots of experiments, we find that the new algorithm plays a good performance in thinning quality, and at the same time, its thinning speed is 3 to 6 times faster than others, so it is a better thinning algorithm.关键词：thinning;fingerprint;image processing;template matching2712|214|0更新时间：2024-05-07
- 摘要：An improved spatial relation constrained FCM algorithm is developed in this paper,the spatial neighbor relation between both pixels and between pixel and regions are defined,and the spatial relation matrix between pixel and regions has been constructed.This matrix is constrained to the partition matrix of the classical fuzzy C-Means clustering(FCM) algorithms and the spatial relation constrained FCM algorithm is formed.This algorithm can automatically segment images with fewer parameters comparing to other algorithms of this category.Many experiments are conducted on the avenue images in which the road intensity is inconsistent,and the results show that our segmentation algorithm has better performance to object's consistency for road images with shadows.关键词：image segmentation;fuzzy clustering;spatial relation constraint3650|251|0更新时间：2024-05-07
- 摘要：The EzStego algorithm based on the palette images is studied, and the well known χ2and RS are analyzed based on whichwe pointed out the irdisadvantages when they are used to detect the lower embedding rate images processed by EzStego. Color pairs and sequential embedding are two characters of EzStego, according to which a new method is proposed. Experimental results show that this method works well for detecting stego images of Ezstego and can achieve better performance comparing to both χ2 and RS in case of detecting little embedded stego images.关键词：steganalysis;information hiding;EzStego;frequency statistic4033|205|0更新时间：2024-05-07
- 摘要：Considering the coherence among neighborhood pixels in an image, a kind of spatial domain watermarking scheme based on support vector machine is proposed.It uses support vector machine to embed the watermark and gains satisfied results.Due to the good learning ability and generalization ability of SVM with limited training samples,it can learn the relationship between the selected pixel and its neighboring pixels well with support vector regression.Then,a bit of the watermark is embedded or extracted by adjusting the value between the selected pixel(i.e.desired output) and the actual output of the trained SVR.Experimental results show that the proposed algorithm has good image perceptual quality and high watermark robustness to common image processing operation and the JPEG compression,which also possesses good security and practicability.关键词：digital watermarking;robustness;support vector machine(SVM);support vector regression(SVR)2741|397|0更新时间：2024-05-07
- 摘要：With the increase of Internet bandwidth, applications based on video stream have been developed rapidly. But the applications, such as digital video, videomail or videophone, are very vulnerable in network environment. Therefore video security has become an important research field. In recent years, a lot of image and video encryption algorithms and techniques have been used to solve the problem. However, most of them have drawbacks in security or speed. Combining with other algorithms’virtues, this article puts forward amethod that uses both stream ciphers to generate pseudo random numbers and block ciphers to do permutation. A 4 dimensional hyperchaotic function is chosen as stream ciphers, and Baker map as block ciphers. The improved Bakermap canwithstand known plain text attack in simple form. And the 4 dimensional hyperchaotic function can not only enhance encryption speed but also increase cipher complexity. This method, which can be simply realized by software, is independent of any kind of video compression algorithms, while providing high security for real time digital video with fast encryption speed.关键词：Baker map;hyperchaotic;video stream;encryption;stream cipher;block cipher2928|239|0更新时间：2024-05-07
- 摘要：Integrative management of massive spatio-temporal data and the relevant capabilities for time series analysis are important research targets of new generation GIS software architecture.At present,access efficiency of spatial and spatio-temporal data based on seamless massive table mechanism needs solid improvement.For managing massive spatio-temporal data effectively and improving spatio-temporal search performance,partitioning and clustering method of large object-relational database platform are extended to spatio-temporal dimension,and performance optimization method for massive spaito-temporal data based on spatio-temporal partition clustering is presented,which is based on extended relational spatio-temporal data model.The search efficiency test on single table of 2GB to 60GB shows that STPC provides better spaito-temporal search performance(about 10.1%) than the normal data organization methods.关键词：massive spatio-temporal data;spatio-temporal partition;spaito-temporal clustering;spatio-temporal partition clustering3641|344|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