最新刊期
卷 17 ， 期 1 ， 2012
- 摘要：In the past century,the Shannon sampling theorem has underlain nearly all the modern signal acquisition techniques.It claims that the sampling rate must be at least twice the maximum frequency present in the signal.One inherent disadvantage of the theorem,however,is the large number of data samples particularly in the case of special-purpose applications.The sampling data have to be compressed for efficient storage,transmission and processing.Recently,Candès reported a novel sampling theory called compressed sensing,also known as compressive sampling (CS).The theory asserts that one can recover signals and images from far fewer samples or measurements,not strictly speaking,as long as one adheres to two basic principles:sparsity and incoherence,or sparsity and restricted isometry property.The aim of this article is to survey the advances and perspectives of the CS theory,including the design of sparse dictionaries,the design of measurement matrices,the design of sparse reconstruction algorithms,and our proposal of several important problems to be studied.关键词：compressed sensing;sparse approximation;incoherence;measurement matrix;sparse optimization7754|444|0更新时间：2024-05-08
- 摘要：Reversible data hiding has extensive applications in fields like medical data management and forensic material authentication.A new reversible data hiding algorithm using AVS video as cover is proposed in this paper.The secret information is embedded into the cover video by simultaneously modifying quantization parameter and quantized DCT coefficients of a macroblock.The original cover video can be completely recovered after the secret information is extracted,and moreover,the data embedding process brings no visual distortion to the cover video.In order to restrain the video bitrate increment caused by this algorithm,an improved reverse zerorun length(IRZL) coding strategy is applied.The experiment results indicate that the new strategy is quite effective in restraining bitrate increment.Under situation that embedding rate is large enough,IRZL coding is more effective than original reverse zero run length(RZL).关键词：reversible data hiding;AVS video;complete video quality-preserving;reverse zerorun length (RZL) coding3635|242|0更新时间：2024-05-08
- 摘要：Recently a number of algorithms have been proposed to overcome the data insufficiency for certain compressible signals.Most of the existing methods are appropriate for piecewise smooth objects,but do not behave very well on texture-rich object.In this paper,a new optimization model for tomography reconstruction for texture-rich object is proposed based on morphological components analysis,in which compound regularizations are exploited.Furthermore,an alternating iterative algorithm is presented to solve the relevant optimization problem.We compare its numerical performance with two recent algorithms,which demonstrate that the proposed method is highly efficient especially in preserving texture features.关键词：magnetic resonance imaging;rphological component analysis;alternating iterative algorithm3231|164|0更新时间：2024-05-08
- 摘要：It is important to extract the information from images which have poor contrasts and colors acquired in bad weather.This paper proposes an effective method to remove haze from a single input image. First, focusing on sea images with deep and wild view, the formulas of relative depth are given based on a projection model from 3D to 2D. Then, using the dark channel prior, the sky intensity and reference depth can be obtained more accurate. At last, the degrade images can be recovered by relative scene depths model and atmosphere physical model.Our results demonstrate that our approach performs better than the dark channel prior method.关键词：image dehazing;reference depth;dark channel prior;sky intensity3797|239|0更新时间：2024-05-08
- 摘要：To estimate a range of image distortions,a novel no-reference image quality assessment method is proposed based on wavelet multi-scale transformation.For natural scene statistics (NSS) model,the sub-band energy of wavelet transformation has a linear distribution with scale index.According to this principle,the energy distribution of ideal image could be predicted from high-scale sub-band energy,which was not badly affected by distortion.Meanwhile, an effective method for identifying and compensating for an inappropriate distortion was presented.Finally,the quality metric was constructed by quantifying the difference between predicted energy and real energy in degradation image.Experimental results showed that the new method was consistent with subjective assessment and outperformed the other methods.关键词：no-reference image quality assessment;natural scene statistics (NSS) model;wavelet transform;multi-scale prediction;energy compensation3937|204|0更新时间：2024-05-08
- 摘要：In this paper, we present a new algorithm to remove Gaussian noise in images by combining region segmentation and bilateral filtering.The pixel-based bilateral filter suffers difficulties in an appropriate selection of the smoothing parameter due to the disturbance of noise,which poses obstacles to the filtering process in dealing with edges and hence degrades the preservation of image structures.Based on the segmentation of images,this work takes advantage of the region map to facilitate the filtering process.Specifically,noise characteristics within regions and similarities between regions are utilized to derive smoothing parameters corresponding to the involved pixels.By applying filtering to pixels within regions and between regions in different means,the algorithm gains enhanced adaptability to image structures.Relative to some existing bilateral filter algorithms,testing results have demonstrated that the proposed algorithm improved the de-noising performance especially on the preservation of structure information.关键词：bilateral filtering;structure preserve capability;region adjacency graph;region similarity4098|223|0更新时间：2024-05-08
- 摘要：In this paper,a new method is proposed for removing and restoring random-valued impulse noise in images in order to preserve fine details.At first,based on a similar neighbor criterion,seed pixels are extracted.Then switch median filter is applied to the seed pixels for removing the misjudged seed pixels with an adaptive window size.Lastly,the missing seed pixels are probed firstly and then the noise pixels are restored both using the seed pixels.Compared with other well-known methods,this technique achieves superior performance.关键词：image denoising;random-valued impulse noise;noise ratio;noise detection2977|172|0更新时间：2024-05-08
- 摘要：RDO (rate-distortion optimization) plays an important role in video coding systems and has a great effect on coding efficiency.The most widely used RDO strategy uses MSE or other similar metrics for distortion modeling,which is not a good metric for subjective evaluation.In order to improve the perceptual quality,a novel perceptual distortion model is firstly proposed which takes the perceptual properties of texture and luminance into consideration.Based on the perceptual distortion model,a TL-RDO (texture and luminance based RDO) strategy is proposed which adjusts the Lagrangian multiplier dynamically according to visual perception.The simulation result shows that TL-RDO gets higher coding efficiency than the famous QP-RDO.Moreover,it has low computational consumption compared to other perceptual RDO strategies,which is suitable for real-time systems.关键词：perceptual quality;perceptual distortion model;video coding;HVS;H.264/AVC;RDO3122|132|0更新时间：2024-05-08
- 摘要：A novel single-channel blind separation algorithm for permuted defocus blurred image is proposed based on parameter estimation in this paper.The defocus blur radius is estimated by the characteristics of permuted image in the frequency domain,and then the permuted image is restored by performing the Lucy-Richardson(L-R)blind restoration method.The ringing effect of restored image is measured by defining the sum of absolute pixel gradient,and the permutation mixing matrixes can be accurately estimated by classifying the ringing effect of each sub-block,thereby separating the source images.Simulation results show that the proposed algorithm is able to achieve better separation efficiency for the permuted defocus blurred image with various permutation operations and a better performance on the robustness against Gaussian noise and lossy JPEG compression.关键词：permuted image;blind source separation;defocus blur;parameter-estimation;single-channel3414|121|0更新时间：2024-05-08
- 摘要：Speckle noise is the major factor affecting the quality of side-scan sonar images. Side-scan sonar image despeckling has therefore a great significance to object identification and image processing.According to side-scan sonar image characteristic and distribution of speckle noise,a despeckling method based on block-matching and 3D filtering (BM3D) is proposed in this paper.A Rayleigh distributed multiplicative speckle noise model is established according to a seabed scattering model.Rayleigh distributed of speckle noise is changed into Gaussian distribution by the Gaussian smooth function power transform.The multiplicative noise is changed into additive noise by a logarithmic transform.Then,the transformed images are filtered by an efficient BM3D algorithm.Last,the final images are achieved by inverse transform.The experimental results show that the new algorithm has a better performance in terms of edge preserving and denoising than other spatial filtering,wavelet and Curvelet domain filtering methods.关键词：side-scan sonar image;speckle noise;despeckling;block-matching and 3D filtering3478|129|0更新时间：2024-05-08
- 摘要：A posed model and a new diffusion method are introduced. After analyzing the diffusion coefficient function of the Perona&Malik(PM) model, We found that it is too sensitive at edges,which is the main reason for the PM model’s ill-posing.Doing some amendments to the function, we got a posed anisotropic diffusion model. Our new model has a better form for double diffusions and filter factors,which the ill-posed models don’t have. We divided the model into two different modules:a smoothing module and a denoising module. The two modules separate the operations of smoothing and denoising. As a result,Our new method can improve the impression of smoothness and reduce the noise remove.Our experiments on real images show that the proposed algorithm outperforms many typical image diffusion algorithms on visual effects.关键词：partial differential equation(PDE);smoothing and denoising;direction of the image feature;anisotropic diffusion3480|125|0更新时间：2024-05-08
- 摘要：An approach to recognize types of fascicular groups from nerve slice image by the gray level multi-direction gradient and its 2nd derivative gradient is proposed in this paper. First, some pixels are selected arbitrarily from a slice image and the gray level multi-direction gradient and the 2nd derivative gradient of their neighborhood areas are calculated. Then the frequency and amplitude of the multi-direction gradient and 2nd derivative gradient curves are extracted as the texture features of the arbitrarily selected pixels neighborhood. Second, the algorithm for recognizing the types of fascicular groups in nerve slice image is proposed based on the improved rough K-means clustering.The parameters influencing the recognition results are analyzed. The experimental results show that the approach can not only classify the types of fascicular groups accurately but the recognizing results are unrelated with the parameters, which reflect its good adaptivity.关键词：slice image;texture;multi-direction gradient;nerve fascicular groups3202|120|0更新时间：2024-05-08
- 摘要：Multiple-target tracking in complex scenes is one of the most complicated problems in computer vision. Handling occlusions between objects is the key issue in multiple-target tracking. This paper introduces a method of motion segmentation into the object tracking system, and presents a SPA (skeleton points assign) based occlusion segmentation approach to track multiple people through complex situations which are captured by static monocular cameras. In the proposed method, we select the skeleton points and evaluate their occlusion states by bottom information like optical flow; then we assign these points to different objects using advanced semantic information, such as appearance, motion,and color.Finally a dense classification of foreground pixels is used to accomplish occlusion segmentation. Object tracking is handled by a particle filter-based tracking framework, and a probabilistic appearance model is used to find the best particle. Experiments are performed using the public challenging data set PETS 2009. The experimental results show that our approach can improve the performance of the existing tracking approach and handle dynamic occlusions better.关键词：multiple-target tracking;occlusion segmentation;particle filter;skeleton points;motion segmentation3172|182|0更新时间：2024-05-08
- 摘要：In order to improve the stability and discrimination of local feature combination for image representation,two image mediate-level representations,Inter-CSP(inter-class common stable pattern) and ntra-SSP(intra-class special stable pattern) are proposed.The details of processing are given,which can be divided into statistic-filtering,pattern decomposition,pattern summarization,and item-based geometric relation modeling on frequent item_sets mined from image semi-local features.A recognition framework is introduced based on Inter-CSP and Intra-SSP.The experiment results demonstrate that these two kinds of patterns are superior to classical methods.关键词：frequent item set;pattern decomposition;pattern summarization;stable pattern3172|205|0更新时间：2024-05-08
- 摘要：In this paper,we present a two-stage vessel recognition method based on structural features of high resolution TerraSAR-X images.We analyze two categories of vessels:tanker and cargo ship.The tanker class includes tube-showing-tanker and no-tube-showing-tanker.The cargo ship class includes closed rail cargo ship and open rail cargo ship.In the first recognition phase we distinguish closed rail cargo ships and tube-showing-tankers as one class from no-tube-showing-tankers and open rail cargo ships based on whether the ship body possesses holes.The second phase discriminates closed rail cargo ship from tube-showing-tanker based on whether the longest lines of the ship exist in the middle region of the vessel or not.The discrimination of open rail cargo ship from no-tube-showing-tanker is based on whether or not the ship body could cut off line.We obtained good experiment result with our method.关键词：high resolution;TerraSAR-X;vessel recognition;structure feature;SAR3319|190|0更新时间：2024-05-08
- 摘要：An image-matching method for robot environmental perception based on hybrid features from color histograms based on the scale-invariant feature transform (SIFT) is proposed.The SIFT is combined with color histograms to make a compromise between high perception accuracy and real-time processing needs. First,images are processed by making an average of the lightness,then the extracted features are added to the main color histogram,which is more robust against lightness and dynamics in the environment.The number of SIFT values is controlled and local color statistical information is added to the SIFT,which is more accurate and faster for real-time matching.After wards,the process of features-retrieval is accelerated by hierarchical matching.Finally,the scheme is optimized using the proposed reasoning method based on previous knowledge from databases,to further improve the accuracy of perception the simulation and experiment results show that when the scale of the database is growing,the advantage of the proposea method proposed is prominent.关键词：vision-based environmental perception;color histogram modifying;SIFT modifying;knowledge reasoning;algorithm optimization3438|183|0更新时间：2024-05-08
- 摘要：The computation cost of the current exemplar-based image inpainting algorithms is high due to the fact that the current algorithms compare the exemplar patches with the inpainting patch one after another.In this paper we propose a fast image inpainting algorithm based on color texture distribution analysis which determines the range of exemplar sets according to local texture variances,thus avoiding wasting amount of time searching and comparing with exemplar patches in vain when the examplar patches set is too large and the lack of the diversity when the examplar set is too small.The results show that the presented algorithm can ensure good performance of visually high quality while being consistent with the surrounding textures and greatly improving the computation efficiency greatly.关键词：image inpainting;image completion;texture synthesis;seamless3430|202|0更新时间：2024-05-08
- 摘要：Anti-aliasing for circles is complex and the existing algorithms are not working satisfactory.To improve the efficiency and the anti-aliasing effect,an integral algorithm is presented for circle anti-aliasing based on midpoint generating algorithm.The new algorithm finds the grayscales of each pixel according to the distance between the center of the pixel and the circle.A circle with a 64 levels grayscale can be drawn.The algorithm abandons the two-order epsilon and corrects the error by simple calculations.We built a method to forecast the grayscale change between neighboring pixels using integer shift,addition,and comparing without using floating-point and divisions,making it easy to implement the algorithm on hardware.The results show that the anti-aliasing effect and its efficiency have been improved.关键词：circle;anti-aliasing;integer operation;grayscale4703|178|0更新时间：2024-05-08
- 摘要：A kernel principle component analysis method based on tensor algebra is proposed for feature extraction.It can reduce the huge computation cost due to increasing dimensions,while considering the information of known classes.First the kernel principle component analysis method is applied to each class of targets to build their corresponding feature spaces.Then,the collection of feature spaces is unified into a higher dimensional space after introducing the operation of the tensor product.Hence,a linear principle component analysis method can be directly applied on this feature space in order to construct the proper feature space to both reflect the characters of each class and lower the cost of computation.The recognition experiments showed that the cost of computation and memory can be decreased heavily compared to the approach that builds the feature space by using the kernel principle component analysis method directly.关键词：kernel principle component analysis (KPCA);tensor algebra;automatic target recognition;SAR4502|194|0更新时间：2024-05-08
- 摘要：A segmentation method for high-spatial resolution remotely sensed images segmentation based on the hierarchical representation of the image content is proposed in this paper.The hierarchical representation is built through a hierarchical merging process based on an initial segmentation.First,we build the region adjacency graph (RAG) on which a Markov random field (MRF) is defined.Then the hierarchical merging process,where the merging criterion includes multi-spectral,edge and shape features,is applied and recorded on the RAG to obtain the hierarchical representation.It is quite efficient to produce a number of segmentation results with different precisions using the relationship of different objects in the hierarchical representation.Experimental results on QuickBird images show that the proposed method can greatly improve the segmentation efficiency and produce high quality results.关键词：remotely sensed imagery segmentation;hierarchical representation of image content;hierarchical merging;Markov random field;region adjacency graph3713|184|0更新时间：2024-05-08
- 摘要：In this paper we propose a grid simplifying method for road networks which preserves the structural patterns.The proposed approach firstly identifies the grid structure,and then merges grids contained by other grids and aggregates these grids into clusters according to their connectivity.The proposed method simplifies grids according to the principle of information minimum loss that is characterized by the grid index,the principal direction,and the elongation index.Two road datasets of Zurich and Wuhan were selected to evaluate the validity of the proposed method.Experimental results show that this method well preserves grid-like pattern.The simplified results provide a good data source for visualizations on small screen devices,multiple representation,and progressive transmission.关键词：road pattern;grid simplification;information loss3726|150|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