最新刊期
卷 17 ， 期 10 ， 2012
- 摘要：Many tasks in image processing,computer vision and computer graphics require image interpolation or resampling in order to obtain data at locations that do not coincide with the grid points where the digital image values are known. While image interpolation is fairly well understood for scalar images,not much research has been done so far with respect to the interpolation of tensor fields. In this paper,we present a systematic review of the current research status regarding tensor valued image interpolation. Existing approaches of tensor valued image interpolation are fully analyzed and categorized according to their mathematical framework. First,the drawbacks of tensor calculation under the Euclidean framework are pointed out. Then,the transition of research efforts from under Euclidean framework towards under the Riemannian framework are sorted out. After that,different evaluation metrics of tensor field interpolation methods are compared. Finally,future research directions are discussed.关键词：tensor valued images;interpolation method;Riemannian metric;tensor component2735|110|0更新时间：2024-05-08
- 摘要：In order to provide a large,scalable embedding capacity and maintain a good visual quality of the stego-image,In this paper, a secure steganography based on modulo and its cyclical characteristic is presented. First,the secret data stream is converted into an -ary notational system data stream. Then,according to the characteristic of the remainder cycle,one -ary notational information will be embedded into one pixel of the cover image by a modulus function with security features,until the secret data is embedded completely. Different parameter n gets different rates of embedding and visual quality of the stego-image. Theoretical analysis and experimental results show that the algorithm can maintain a high embedding rate while maintaining good visual quality of stego-image,and good security. The information-embedding rate is highly flexible,and the method can choose the steganographic scheme according to the actual requirements,for which the embedding rate and stego-image quality are ideal. It is more practical than other similar algorithns.关键词：digital image;modulus function;remainder cycle;steganography2470|49|0更新时间：2024-05-08
- 摘要：In this paper,we review the past 15 years of research in the field of night vision multi-sensor image coloration (render night vision image in color) and reveal the general coloration model. On this basis,a new coloration method using fast multi-modulation fusion (FMMF) and color transfer is designed for low-light and infrared image pairs. The coloration process is based on YCbCr color space. First,the fused image uses a fast multi-modulation fusion to merge the source images information to be assigned to the Y channel;then,the Cb and Cr channel is combined using Toet’s method,which extracts the common component from the source images. Finally,the false-color image is obtained by using color transfer technology to the prior pseudo-color YCbCr image. Experiments show that the result of our method is information that is more salient,has a higher color contrast,and a more natural color appearance than others. Due to the use of fast multi-modulation fusion,the coloration process is efficient and the parameters are adaptive,and our method meets the real-time applications.关键词：image fusion;night vision;coloration;color transfer2905|53|0更新时间：2024-05-08
- 摘要：Inspired by Weber’s Law,we propose an adaptive method to build up scale-space based on "just noticeable difference". This is because human perception of patterns depend not only on the change of a stimulus but also on the original intensity of the stimulus. According to this point of view,we use the ramp edge and step edge to calculate the image information amount based on Marr’s theory. The information lost between the adjacent scale layers is then gained in experiments when the human vision can feel the "just noticeable difference" . Finally,the scale-space is constructed adaptively by curve fitting method. The experimental results show that the algorithm can exhibit the characteristics of human visual perception. In the matching experiments,the number of matched points can be increased at least by 25% by the adaptive algorithm and it impressively outperforms the other scale-space construction method in denoising field.关键词：scale-space;information amount;visual perception;just noticeable difference (JND)2748|46|0更新时间：2024-05-08
- 摘要：The weight function of the non-local means denoising method has a certain degree of irrationality at edges,in which cannot distinguish between the denoising roles of the patches at the two sides of an edge. However,when the center pixel of the patch gets more attention,the different denoising roles can be measured. In the light of the bilateral filtering method,the weight function of the non-local means method is revised. Experimental results on several images show that our new method greatly outperforms the classical non-local means method,and it has certain advantages over the latest improved non-local means method.关键词：image denoising;non-local means;gauss white noise;bilateral filtering4513|110|0更新时间：2024-05-08
- 摘要：We know that Principal Components Analysis(PCA) can represent each face image in terms of a linear combination of the eigenface, we also know that the PCA algorithm gives the best representation of images under the sense of minimum mean square error. However,PCA only compares the Euclidean distance between projection coefficients of samples and ignores the residue between the original sample and its reconstructed one. Therefore a new concept called dissimilarity distance metric is proposed in this paper. We project the two images into the same subspace and then characterize the similarity between pairs of samples by comparing to both the projecting coefficients and the approximation errors simultaneously. The higher is the value,the more dissimilar are the two samples. Different from Locality Preserving Projections,a new method,called Dissimilarity Preserving Projections,uses the concept of the dissimilarity above,and constructs the dissimilarity scatter matrix. This algorithm does not have to pre-set the number of neighbors,finally it gets the optimal projection subspace by maximizing the Objective function. The experimental results on AR and FERET face image database demonstrate the effectiveness of the proposed method.关键词：face recognition;dissimilarity distance metric;feature extraction;residue;projecting coefficients2589|56|0更新时间：2024-05-08
- 摘要：Non-orthogonal Binary Subspace (NBS) is an essential image representation method,whose major advantage is to support high-efficiency image matching. However,generating NBS is itself time-consuming which makes the method difficult to be applied in real-time applications. In this paper,a parallel template-generating algorithm based on Compute Unified Device Architecture (CUDA) is proposed. The generating procedures are divided into three phases in our approach. For each phase,parallel task distribution is used to fully utilize the capability of the GPU. Experimental results demonstrate that our algorithm achieves 60~120 times speed-up,compared to the same template-generating task on the CPU. Additionally,we apply our algorithm on the template-matching problem,and observe significant improvement in efficiency.关键词：non-orthogonal binary subspace;Haar base;matching pursuit;template representation;template matching;compute unified device architecture (CUDA)2643|34|0更新时间：2024-05-08
- 摘要：In this paper,we present a general framework to discover image categories automatically.The algorithm includes two parts:1)we pose the problem of category discovery as an automated graph partition task. Each graph vertex indicates an image,and a partitioned sub-graph consisting of connected graph vertices representing a clustered category. The model of each image category can be learned by stepwise feature selection using the Adaboost algorithm. 2)A MCMC-based stochastic algorithm,the Swendsen-Wang Cuts (SWC),is adopted to solve the graph partition fast. Compared to traditional random cluster sampling techniques,SWC converges faster. We apply our method on two image datasets,and the experimental results demonstrate superior performance of our method over other popular state-of-the-arts methods,including Kmeans,pLSA,and Affinity Propagation.关键词：unsupervised image categorization;AdaBoost;graph partition;Swendsen-Wang cuts(SWC)3125|79|0更新时间：2024-05-08
- 摘要：In order to solve the problem that the current approaches cannot suppress the background clutters effectively,which results in a poor detection performance,a new infrared dim target detection approach is presented,which is based on background suppression by artificial immune network (aiNet) and threshold segmentation by k-means cluster of rows and columns. First,the aiNet is combined with Robinson guard to build the adaptive local spatial background models as fuzzy topological memory antibody bank. In the process of antibody bank modeling,a series of antibody evolution strategies are designed based on self-organizing maps (SOM). With these models,background clutters are suppressed according to the degree of fuzzy match between pixels and models. Then,the proposed adaptive segmentation algorithm based on k-means cluster of rows and columns is used to detect the true targets. Experimental results show that the F1 measurement of the proposed approach is up to 99%. The proposed approach is able to build the spatial background models adaptively according to the local change of image,and suppress the background clutters and highlight the targets effectively. It is capable of improving the signal-to-noise ratio of images and detecting targets effectively.关键词：dim target detection;artificial immune network;self-organizing map;Robinson guard;k-means cluster of rows and columns3186|57|0更新时间：2024-05-08
- 摘要：Inspired by Weber’s Law and the biological model of synergistic center-surround receptive field,we propose a center-surround hypothesis for image saliency detection and design a center-surround structure with circular topology. Based on this model,a saliency detection method fusing local and global features is proposed. It extracts the center-surround structure of each pixel,using Weber’s Law based gradient orientation to represent local saliency,and using the relative intensity differences of the center region against the overall mean to represent global saliency,then it gives the final saliency map by linear combination of local and global saliencies. Comparison experiments and precision-recall curve demonstrate this detector has better performance,and has strong response in active region while inhibits other regions.关键词：visual saliency;Weber’s law;center-surround structure;local saliency;global saliency3319|66|0更新时间：2024-05-08
- 摘要：In this paper we describe the use of image processing methods to extract important information from meteorological facsimile maps for analysis and integration with other information. We propose a method based on matching feature points to judge intersecting lines. Furthermore,we propose a method based on vector products to recognize dotted line segments on vorticity lines,and we combine these methods to extract triangles on frontal lines. Experiments show that these methods work effectively.关键词：meteorological facsimile maps;feature points;geopotential height lines;vector product;vorticity lines3267|54|0更新时间：2024-05-08
- 摘要：A spatial scene is a set of objects along with their spatial and,non-spatial attributions and their mutual spatial relationships. In this study,we propose a spatial scene-matching model considering the object area,attribution,topology,and orientation relationships. We first build a scene semantic description model by combining the "gradual change" model and attributed relationship graph (ARGs). We then construct our scene-matching model that evaluates the scene similarity by the self-proposed similarity measures. This method objectively describes semantics of spatial,especially geographic scenes,and performances well on spatial scene matching. It has a good prospect in intelligent retrieval of spatial data.关键词：spatial scene;spatial relationship;topology;orientation;similarity matching;remote sensing image retrieval3821|73|0更新时间：2024-05-08
- 摘要：Delaunay triangulation will play an important role in numerical simulation of the geophysical process in the future. The divide-conquer algorithm is one of the well-known traditional fast Delaunay algorithms,but its complicatied merging procedure has limited the applicability of this algorithm. In this study,we propose the concept of universal operators,which are used to simplify the construction of the Delaunay triangulation algorithm. Several operators are extracted from some previous Delaunay algorithms,and three new operators are used in the merging process of the divide-conquer algorithm. The operator of expanding a triangle is developed for constructing each new triangle and for managing the topology information and the linked list which represents the border edges of the triangulation. The operator of filling basins is developed for automatically filling the basins by creating new triangles using recursions. The triangulation-merging operator is developed for linking two sub-triangulations using one new triangle,transforming the two original linked list of border edges into one new linked lists with a new sequential order,and filling the two basins by calling the operator of filling basins. All these operators are based on the same set of data structures and the linked list represented triangulation hull. Using operations on the linked list,all search operations in the triangulation hull are avoided,which make the final algorithm very concise and fast. The operators are successfully used for construct other Delaunay algorithms besides the divide-conquer algorithm. Large point datasets generated stochastically as well as LiDAR point clouds are used for testing the operator-based algorithms,and the result shows that the algorithms can correctly generate triangulations. Meanwhile,it is shown that the divide-conquer algorithm based on the operators are almost as fast as the horizontal expanding algorithm.关键词：Delaunay triangulation;universal operators;merging of triangulations;basin-filling;divide-conquer algorithm3595|58|0更新时间：2024-05-08
- 摘要：In order to overcome the disadvantage of being sensitive to model gesture and noise in the present mesh segmentation algorithms,we present a consistent mesh segmentation algorithm based on Laplace spectral embedding and Mean Shift. We convert mesh into a normal form from the space domain to the spectral domain by using the Laplace-Beltrami operator. The noise is suppressed and spectral embedding enhances the structural segmentability. We adopt Mean Shift,a nonparametric kernel clustering technique,to gain the visual meaningful semantic patch or sub-mesh in the spectral domain. The experiment results show that the proposed algorithm can yield meaningful result rapidly and effectively for meshes which has an evident branch structure.Meanwhile,this approach is invariant to pose of model and robust to noise.关键词：Laplace spectral embedding;Mean Shift;consistence segmentation;triangle mesh3276|56|0更新时间：2024-05-08
- 摘要：In this paper,we present a system for automatic reassembly of broken 3D solids based on fractured surfaces matching. First,the fragments are segmented into a set of surfaces bounded by edge using a region growing strategy according to volume integral invariants,and then these surfaces are classified into the original surfaces and fractured surfaces by computing their perturbation of the normal vectors. Second,a small number of salient matching point pairs are obtained through first comparing feature point’s volume integral invariants and then comparing the feature point’s neighbor regions based on compatibility constraint. After that,we use exhaustive search method constrained by triangle similarity and voting scheme to match the fractured surfaces. Finally,we employ a sub-graph merging algorithm based on backtracking to merge all matched fragments until the object is reassembled. Experimental results show the algorithm is capable of reassembly of broken 3D solids.关键词：reassembling fractured objects;integral invariants;fractured surfaces segmentation;fractured surfaces matching;fragments reassembly3107|36|0更新时间：2024-05-08
- 摘要：Based on real nuclear explosions,a comprehensive explosion model is developed,which can achieve real-time simulation of nuclear explosion. Spherical expansion is adopted to express the shock wave,Additionally,fast collision detection is realized. The splash of the nuclear explosion is represented by using particle dispersion within the view range. The Navier-Stokes equations are simplified and solution strategies are optimized.Furthermore,the mushroom is realized with complex forms motion. Experimental results show that using the model can achieve a real-time simulation of the effects of a nuclear explosion. By controlling and adjusting the parameters,many different types of nuclear explosion effects can be obtained.关键词：nuclear explosion;mushroom;Navier-Stokes equations;real-time3094|54|0更新时间：2024-05-08
- 摘要：In order to deal with the limitation of probabilistic tractography which may produce diffuse results that suggest connections in unexpected regions,we propose a fast and novel global fiber tracking method for DTI (Diffusion Tensor Imaging)data. Our method is inspired by the ant colony optimization technique,which considers both:the local fiber orientation distribution and the global fiber path in a collaborative manner. We first construct a global optimization model that captures both global fiber-path and the uncertainties in local fiber orientation between two regions. The local fiber-orientation density function with uncertainty is modeled with a Bayesian approach. Then,an ant colony global fiber-tracking algorithm is presented using a new learning strategy where the probability associated with a fiber is iteratively maximized. In the algorithm,the pheromone model is constructed using the von Mises-Fisher function and the ant fiber tracking technique based on pheromone model is developed. Finally,the proposed algorithm is validated and compared to alternative methods using both synthetic and real data.关键词：diffusion tensor imaging;swarm tracking;tractography;global fiber tracking3055|61|0更新时间：2024-05-08
- 摘要：In this paper,we propose a new method that uses dual quaternion as the mathematical tool for geo-positioning remote sensing images of linear-array CCD. Dual quaternion is used to establish an universal imaging model of remote sensors and to define the position and attitude of the light beam. In order to overcome the strong correlation between the imaging geometry parameters (exterior orientation),we scan the spiral movement of the sensor light and achieve the transformation between image points and ground points. According to the theory of linear skinning blending of the rigid body transformation,we decompose the rigid body transformation matrix into two parts:translation and rotation. In order to calculate the exterior orientation elements of the remote sensing images of linear-array CCD,we use linear interpolation to the translation part and spherical interpolation to the rotation part. According to the established imaging model,geo-positioning experiments are made with Geoeye-1 remote sensing images. The results show that the new algorithm can obtain better geo-positioning accuracy than traditional algorithms,and succeeds in solving the correlation problem between the positioning parameters.关键词：geo-positioning;dual quaternion;line-array CCD images;imaging model;space resection3192|58|0更新时间：2024-05-08
- 摘要：Geolocation accuracy is an important factor for remote sensing applications. In order to improve the geolocation accuracy of FY-3B/MERSI (medium resolution spectrum imager) data, we analyze how the installation matrix impacts the accuracy of the geolocation results. When the instrument coordinate system rotates around the X-axis, Y-axis and Z-axis, the remote sensing images will shift left or right, up or down, and they will rotate. According to the relationship of ground control points and land-sea mask to estimate the error of the installation matrix, the accuracy of the image geolocation results can be about one pixel. The geolocation results from the operate system updated the installation matrix are stable and reliable. The new matrix includes not only the relationship between the installation of the instrument and the satellite platform, but also the systematic bias of the internal device due to vibration, environmental changes and measurement errors of the satellite’s position, velocity, and attitude.关键词：geolocation;installation matrix;FY-3B/MERSI;preprocess3822|49|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