最新刊期
卷 12 ， 期 12 ， 2007
- 摘要：Motion estimation refers to estimating 2D motion vector field of the scene or object according to temporal information redundancy in a clipped video.It plays a key role in the super resolution image restoration technique,which maps pixels of all low resolution observations onto corresponding pixels in the reference frame.So its accuracy is highly required.Block-based motion estimation is widely used in super resolution image restoration.This paper introduces the model of block-based motion estimation,summarizes four categories of fast motion estimation methods,describes in detail the search process of several influential algorithms belonging to the first category,and then compares several fast block-matching algorithms.关键词：super resolution image restoration;motion estimation;block-matching algorithm;search pattern3074|150|0更新时间：2024-05-07
- 摘要：Edge detection is a fundamental issue in synthetic aperture radar(SAR)image interpretation.The multiplicative noise makes edge detection in SAR image extremely difficult,and the typical edge detectors based on gradients are inefficient when applied to SAR images.So developing edge detectors especially for SAR image is of great importance.This paper investigates the problem of edge detection for SAR images comprehensively and thoroughly.It first describes the problem of edge detection for SAR images,and then discusses several well-known edge detection methods and two key points in SAR image edge detection-edge thinning and localization.To evaluate edge detector quantitatively,some performance criteria are summarized.In the end of this paper,conclusions are made and the possible future work is explained.关键词：SAR images;edge detection;edge localization4084|233|0更新时间：2024-05-07
- 摘要：Image mosaic has been researched for several years,but still it is hard to realize mosaic without ghosting and exposure difference.In this paper,for resolving these two problems,a new image mosaic method based on graph cut is presented.This method integrates two techniques:first graph cut,it is used to calculate a robust graph cut seam-line for deghosting by initializing the weighted directed graph with gradient direction histogram of each pixel;second Poisson image fusion,to further eliminate the exposure difference by an overlap transition Poisson image fusion procedure.Experimental results demonstrate the efficiency of our proposed method.关键词：image mosaic;graph cut;Poisson image fusion;ghosting;exposure difference3751|194|0更新时间：2024-05-07
- 摘要：Image super-resolution reconstruction(SRR)refers to a signal processing approach which produces high-resolution images from observed multiple low-resolution images.Many image SRR algorithms assume that the blurring process,i.e.,point spread function(PSF)of the imaging system is known prior to reconstruction.However,the blurring process is not known or is known only to within a set of parameters in many practical applications.In this paper,we propose an approach for blind image SRR based on double regularization by parametrizing PSF.A space-adaptive regularization method for image SRR is used to preserve detail at the textured regions and suppress noise in the smooth background.In the scheme,PSF parameter(s)and the high-resolution image are estimated by alternating minimization method.The demand for precision of minimizations is varied during the optimization procedure in order to reduce the computation cost.Experimental results from a synthetic image sequence show that blur parameters are approximated actually and the reconstructed image is visually pleasing.关键词：image super-resolution reconstruction;blind deconvolution;resolution enhancement;point spread function(PSF)estimation3192|159|0更新时间：2024-05-07
- 摘要：The traditional image interpolation involving zeroth-order,bilinear,and cubic spline interpolation mostly depends on a piece-wise continuous and smooth modeling of image.In view of the ability to capture the feature of natural image connected with the modeling,the quality of interpolated image is not desirable.A new interpolation algorithm based on the wavelet-fractal(FW)coding is proposed with more efficient interpolation strategy.By the aid of the local self-similarity modeling of image,the new interpolation algorithm using FW coding is implemented by the dilation of tree wavelet.The super-resolution zeroth subband is optimized predicted by the first subband in the wavelet domain.And then the interpolated image can be obtained by the inverse wavelet transform.The experiments of the standard images suggest that the new algorithm can obtain distinct texture and edge of the images and higher peak signal-to-noise ratio(PSNR)compared with the traditional bilinear strategy.As a result,the image interpolated by the new algorithm is accurate and real.关键词：image interpolation;fractal;wavelet2762|158|0更新时间：2024-05-07
- 摘要：This paper introduces adaptive lifting wavelet transform algorithm firstly.Prediction and update operator have been adjusted adaptively according to information partial characteristic and accurate match to the process information.The algorithm can decrease the amount of calculation and computational complexity in the image wavelet analysis,can carry out the same address operation,and is advantageous to recur to the DSP hardware realization.Secondly,this paper proposes to adopt adaptive depth first search strategy through definition floating threshold,and improved image coding efficiency.The experiment results show that the new image compression scheme has been improved and it is better than traditional algorithm in the aspects of image quality,reduced the image bit number,compression effect and compression coding efficiency.关键词：image compression;adaptive lifting scheme;adaptive depth first search;floating threshold2639|174|0更新时间：2024-05-07
-
A Delaunay Mesh-based Motion Estimation and Compensation Scheme based on Redundant Wavelet Transform
摘要：In this paper,a new motion estimation and motion compensation scheme based on irregular mesh is proposed using a shift invariant redundant wavelet transform.Control points and PMA(potential motion areas)are identified with a simple correlated operator in the redundant wavelet subbands,while motion estimation is completed through incorporating the PMA into block matching algorithm in spatial domain.The motion compensation is realized through an affine transform mapping triangles from one frame to the other.Furthermore,we give a mask of the PMA.We compared it empirically to the two other ME/MC methods both based on regular mesh deployed in the spatial and based on irregular mesh deployed in the redundant wavelet domain.Experimental results indicate that the complexity has been reduced and the compensation effect could be improved compared to the other two ME/MC methods.关键词：motion estimation and compensation;delaunay triangle mesh(DT);redundant discrete wavelet transform;multi-resolution motion analysis;potential motion areas(PMA)2738|163|0更新时间：2024-05-07 - 摘要：The RPC model has recently raised considerable interest in the photogrammetry and remote sensing community.The RPC is a generalized sensor model that is capable of achieving high approximation accuracy.In this paper an algorithm of computation of parameters of RPC model without initial value is presented.Finally we test the algorithm on SPOT-5,CBERS-2,ERS imagery.关键词：RPC model;rigorous sensor mode;satellite imagery;accuracy5873|267|0更新时间：2024-05-07
- 摘要：Vector median filter(VMF)is a kind of traditional highly effective vector filter which is mainly used to remove impulsive noise from color images.But it fails to distinguish thin lines and boundaries from impulsive noise,and usually filter them out.This paper analyzes quaternion rotation theory and its applications to color image processing.Based on this analysis and by imitating Laplacian operators,a new color impulse detector is presented.Combining this color impulse detector with the traditional VMF,a new kind of switching vector median filters is proposed.The experimental results show that,compared with the traditional VMF and some of the representative vector filters and recently developed vector filters,uhe proposed approach,not only effectively preserves the fine details,but also provides better filtering performance.关键词：vector median filters(VMF);quaternion rotation;impulsive noise;color images2766|162|0更新时间：2024-05-07
- 摘要：Firstly,this paper analyzed the relationship between coherence enhancing diffusion and wavelet shrinkage.Then,it interpreted that the interpretation of the coherence enhancing diffusion under wavelet analysis and the equivalency between coherence enhancing diffusion and wavelet shrinkage in image attribute.Finally,it gave an image de-noising algorithm of coherence enhancing diffusion,which used wavelet coefficients to estimate the image edge according to the wavelet's time-frequent analysis function.Experimental results showed that the presented diffusion factor can orient image edge accurately,and the presented algorithm can reduce the noise effectively.关键词：coherence enhancing diffusion;wavelet shrinkage;image de-noising2768|145|0更新时间：2024-05-07
- 摘要：Both nonlinear diffusion and TV method based ROF model are analyzed in this paper,and a basic coupled method for color image is mentioned.In order to remove noise effectively and preserve edges and key details in color image,considering the information of each channel of color image and the advantages of denoising and edges preservation of TV flow and ROF model,a improved channel coupled diffusion model which based on TV flow is proposed,different effects among uncoupled diffusion model,channel coupled diffusion model and improved channel coupled diffusion model are analyzed,and different effects among forward diffusion,TV flow and backward diffusion based on the new model are also discussed.Experimental results show that,1)The improved channel coupled diffusion model is better preserving geometric information such as edges in addition to its effectiveness for image denoising.2)The properties of forward and backward diffusion based on the new model are not changed in color image denoising.关键词：nonlinear diffusion;total variation;ROF model;TV flow;color image coupling3090|143|0更新时间：2024-05-07
- 关键词：image segmentation;weighting fuzzy c-means clustering algorithm;cluster validity function3212|197|0更新时间：2024-05-07
- 摘要：Image segmentation is an important process in image processing.The quality of image segmentation directly affects the following analysis and recognition.Because evolutionary agent has some advantages such as self-fitness non-linear mapping and ability of parallel disposal,this paper presents an autonomous agent-based image segmentation approach.In this approach,a digital image is viewed as a two-dimensional cellular environment in which the agents inhabit and attempt to label homogeneous segments.The agents rely on some reactive behaviors such as breeding and diffusion.The agents that are successful in finding the pixels of a specific homogeneous segment will breed offspring agents inside their neighboring regions.Hence,the offspring agents will become likely to find more homogeneous-segment pixels.In the mean time,the unsuccessful agents will be inactivated without further search in the environment.It can be seen from our experiment in medical chest CT image and brain MRI image that this method can better extract interesting regions.2469|199|0更新时间：2024-05-07
- 摘要：In this paper,the algorithm of global motion estimation is investigated,and a fast global motion estimation algorithm is introduced.The new algorithm is based on nonlinear dense estimation and applies the affine model.In order to reduce the time cost during calculation,three-level pyramid is utilized in the algorithm.It combines outlier detection algorithm based on blocks and feature pixels selection algorithm in the calculation of each pyramid level.The calculation time of global motion estimation is reduced and the accuracy is improved.The experimental results show that the new algorithm is superior to original algorithms in velocity and accuracy.关键词：global motion estimation;parameter model;affine transform;outlier detection;feature pixels selection3158|163|0更新时间：2024-05-07
- 摘要：In this paper,a scene segmentation method utilizing both visual and motion features is presented.Not only the visual similarity but also the motion consistency of shots within a scene is considered in clustering shots into scenes.In addition,we present a method to merge the over-segmented scenes.Experimental results show the effectiveness of our algorithm.关键词：scene segmentation;shot similarity;shot clustering2795|180|0更新时间：2024-05-07
- 摘要：This paper presents the method which improved the efficiency of information extraction based on feature unit of high-resolution remotely sensed image. To improve the precision of image processing, this paper applied image rough-classification based on large scale and precise-segmentation based on different scales. This paper used parallel computing method to improve the speed of image processing. For the data partition method of parallel computing of remotely sensed image, this paper summarized the general data partition methods and gave the general impelmentaiton method of data symmetrical partition method. After the characteristic analysis of the some special of remotely sensed image, this paper gave the mechanism of improving the efficiency of data partition and presented a new scale data asymmetric partition method, and gave the analysis and implementation of the new method. For the image parallel processing based on remotely sensed image database, this paper presented a new data distributing method. The analysis results show that the new methods can improve the efficiency of parallel computing for some special remotely sensed image in the special condition.3044|170|0更新时间：2024-05-07
- 摘要：This paper addressed the image discriminant analysis problem.By constructing the scattering matrices of the image matrices,Foley-Sammon discriminant analysis(FSLDA)are transformed into a bi-objective optimization problem with uncorrelated constraint for improving the speed of feature extraction and the recognition rate.The efficient projection vector is defined and the efficient projection vector can be obtained from deciding the eigenvector corresponding to the eigenvalue of maximum of a generalized eigen-equation.Compared with the other image projection analysis methods,the proposed method has the following properties:(1)the scattering matrices are directly based on image matrices;(2)the efficient projection vectors are statistically uncorrelated;(3)the within scattering matrix is not necessarily invertible and some matrix inversions are not performed.Finally,the proposed method is tested on ORL and NUST603 face databases.The experimental results indicate that the recognition performance of the proposed method is prior to the other methods,and its speed for feature extraction is faster than the above methods.关键词：image projection discriminant analysis;image effective projection vector;image feature extraction;multiobjective optimization;optimality condition;face recognition2909|130|0更新时间：2024-05-07
- 摘要：Fisher linear discriminant analysis(LDA),a well-known feature extraction method,searches for the projection axes on which the data samples from different classes are far from each other while requiring data samples of the same class to be close to each other.Large margin classifier(LMC),also referred as linear support vector machine,de finds a project direction onto which two classes of the samples projected reach maximal margin.With combination of advantages of both LDA and LMC,the paper develops a novel linear projection classfication algorithm,called Fisher large margin linear classifier.The underlying idea is that an optimal discrimiant vector wbest is found along which the samples of high dimensional input space are projected such that the margin is maximized while within-class scatter is kept as small as possible.In addition,relations to other classifiers are explored in theory in this paper.Finally,the proposed method is tested on ORL face database and FERET face database.The experimental results show that the proposed classifier outperforms other linear classifiers.关键词：large margin classifier(LMC);support vector machines(SVM);Fisher linear discriminant analysis;face recognition3537|163|0更新时间：2024-05-07
- 摘要：Palmprint has been demonstrated as one of the most convenient and powerful method in personal verification.This paper proposes a new characteristic called "orientation-texture feature" in palmprint when analyzing its features.First we adopt method to extract the region of interest by tracking the optimal ellipse in the palmprint area.The method is robust to shift,rotation,stretch and other interference.Then we use multichannel-sampling Gabor filters,which simulates certain characteristics of the human visual system,to process and filter our image.The filter we design can automatically adapt to the image via selecting parameters according to our analysis of the width and orientation of the texture.In the filtering stage,we use different filters from different orientations and widths to extract features in palmprint from ridge,wrinkle,and principle line resolutions,and compute its block energy in polar axis by using an optimized ring-orientation projective algorithm.By adopting FCM clustering algorithm,demonstrate that this approach has its advantage of high verification rate in palm print.关键词：palmprint;Gabor filter;entropy correlation coefficient(ECC);energy block analysis;fuzzy C mean(FCM)2847|169|0更新时间：2024-05-07
- 摘要：Obstacle detection is the main components of cross-country intelligent vehicle guidance.Cross-country environments always have changeful illuminations and complicated terrains.The paper presents a new cross-country obstacle detection method based on binocular vision system.First,we calibrated the parameters of the vision system and studied the coordination transform at first to eliminate the influence of terrains.Second the original images were preprocessed by Gaussian filter and contrast-limited adaptive histogram equalization(CLAHE)method to weaken the effect of noise,light and contrast.Harris corners were located with sub-pixel accurate.Third to guarantee the overall system real-time performance,feature-based matching techniques were studied and fundamental matrix was calculated based on random sample consensus(RANSAC).Fourth continuity restrain was studied to eliminate pseudo matching pairs.Finally data interpolation was introduced to build elevation maps.Edge extraction and morphological processing were concerned to accomplish obstacle detection.Experimental results for different conditions are presented in support of the obstacle detection technology.关键词：cross-country intelligent vehicle;binocular vision;contrast-limited adaptive histogram equalization(CLAHE);feature-based matching;obstacle detection3124|184|0更新时间：2024-05-07
- 摘要：In this paper green components are used to separate the crop rows from its soil background images.To determine the detection peaks and verify lines in Hough transform,a powerful tool for lines extraction from images in noisy or degraded environment,the conventional Fisher discriminant criterion function is modified to project the sample points in an accumulator into a variable.This is regarded as an efficient measurement for the density and orientation of the points distributing collinearly.An optimal mathematical model for identifying multi-rows is presented.Experimental results show that the algorithm can efficiently eliminate the effect of the weeds,and its accuracy and robustness are improved compared with the conventional Hough transform.And it is useful for the row-recognition system.关键词：Hough transform;Fisher theory;crop rows recognition;machine vision navigation2837|172|0更新时间：2024-05-07
- 摘要：Handwritten Chinese characters recognition is the difficulty of character recognition.Based on the mechanism of apery imitation,a kind of Chinese characters codes for computer cognition is presented in this paper to apply to the deformation factors of handwritten Chinese characters and to improve the recognition rate of Chinese characters.The configurations of horizontal stroke,upright stroke,left-falling stroke and right-falling stroke are defined in a fuzzy way.Elements groups of Chinese characters are made for machine cognition.Bearable mistakes codes of various categories are given to the elements which are easily confused.Rules for judging stroke sequence are given.36 kinds of subsidiary configurations codes and bearable mistakes codes are constructed.The code principles and multi-template dictionary of Chinese characters which agree with apery imitation are established.10 000 Chinese characters in Xin Hua Dictionary are standardized coded,the rate of repeated codes of which is 0.48%.After testing the recognition on 100 handwritten Chinese characters in the handwritten Chinese character library of HCCORG and NKIM,the recognition rate is 96%.Emulational experimental results show that this kind of coding applies to the deformation of handwritten Chinese characters well and the rates of repeated codes and wrong codes are low.关键词：off-line handwritten Chinese characters recognition;bearable mistakes code;elements groups;stroke sequence;subsidiary configurations2625|154|0更新时间：2024-05-07
- 摘要：Aiming at the magneto-optic image of aircraft rivet,a new automated recognition algorithm based on fuzzy support vector machine(FSVM)is presented in order to inspect the crack of rivet and its direction.The binary image of rivet is obtained by preprocessing the magneto-optic image;the approximate center of rivet is obtained through the threshold method;the star radial vector radiated from the approximate center of rivet is regarded as the characteristic,and the algorithm based on FSVM is used to inspect the crack of rivet and its direction.The kernel parameter and the penalty constant of SVM are optimized using the grid method.And fuzzy multi-class classification method is adopted to avoid refusal classification and false classification.Experiment results show that good effect of defect recognition is achieved using our SVM classifier,and the request of high real time in automated recognition is satisfied.关键词：magneto-optic image;star radial vector;support vector machine(SVM)3429|149|0更新时间：2024-05-07
- 关键词：bad data;smoothing treatment;cubic spline curve2060|136|0更新时间：2024-05-07
- 摘要：Image warping is an important technique in the field of computer animation.This paper presented a method to implement it.In this method,surface interpolation based on Biharmonic spline was used for image spatial mapping.By using inverse mapping and bi-linear interpolation to implement image resampling and combining image's cross-dissolve,we completed image morphing.Due to the flexibility of this method,animators can easily adjust the numbers or positions of feature points to control the detail of warping without other operations.Experiments show that this approach can produce fluid and natural transition image.Besides,every warped image has both stable image boundary and good visibility.关键词：Biharmonic spline;image warping;inverse mapping;bi-linear interpolation;image morphing4426|189|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