最新刊期
卷 11 ， 期 4 ， 2006
- 摘要：With easier acquisition and storage of video data, the problem of how to utilize them effectively becomes increasingly serious. In recent years, enlightened by data mining, more and more researchers over the world show great interests in video mining, which is a new scientific research area deals with the extraction of video semantic info, implicit patterns and knowledge from video data. By using its technology, intelligent video applications would be well developed for assisting people in decision-making. This paper investigates video mining technology and analyzes the studying progress. The concepts, implementations and applications of video mining are summarized and discussed in detail. The challenges in the studying of this field are also put forward.关键词：multimedia data mining;video mining;data mining4973|243|0更新时间：2024-05-08
- 摘要：Image can be encoded effectively by using the spatial distribution characteristic of wavelet coefficients. In order to make full use of the intraband correlation and interband correlation of wavelet coefficients, an effective image coding algorithm is proposed. In the algorithm, subband quadtrees are combined to a scanning tree, and then bitplanes are encoded with the information provided by the scanning tree. The algorithm has better coding performance than EZW and SPIHT for it effectively encodes the image areas which are rich in texture. In addition, the algorithm is also characterized by features of easy realization, fast execution and little memory consumption. So it reduces the difficulty of hardware realization and has good practicability.关键词：image coding;bitplane coding;scanning tree;wavelet transform2750|196|0更新时间：2024-05-08
- 摘要：Generally fast algorithms for motion estimation have low precision. Based on the basic theory and rule of block matching, a new fast motion estimation algorithm which uses maximum grads and more cycle conjugate search algorithm to improve searching approach, is presented. Its speed is comparable to typical fast algorithm and its precision is compariable to the exhaustive searching algorithm, meanwhile it has veracity, simple algorithm path and is easy to be realized. The results of the comparison are also presented.关键词：image coding;image processing;motion estimation2316|190|0更新时间：2024-05-08
- 摘要：In recent years,multiple description coding(MDC) had been proposed to address robust video transmission problem over error-prone networks.In this paper,a macroblock splitting based multiple description scheme is proposed,and the corresponding redundancy control and mismatch control methods are also given.The proposed scheme is easy to be realized and compatible with the video coding standard.Experimental results show that the proposed scheme of multiple description video coding outperforms the optimal DCT coefficient splitting scheme and the MDTC scheme.关键词：multiple description coding;macroblock splitting;robust video transmission;mismatching control2411|194|0更新时间：2024-05-08
- 摘要：In wireless environment without assurance of quality of service, the video stream compressed by conventional video compress standards is vulnerable, especially in intra coded frame. So it is necessary to make modification on structure of conventional video stream in order to enhance its error resilience performance inwireless environment. In this paper, one method is proposed to improve robustness of intra coded frame in wireless transmission, it firstly analyze the sensitivity to error propagation of different segment in video stream when there are channelbit errors. And thenwith this analysis results and respective role of different segments of video stream on reconstruction of intra coded frame, important information for intra coded frame is determined. Finally, the proposed method exploits protection, extraction and aggregation of determined important information in intra coded frame tomitigate the extent of error propagation inside frame reference on traditional data partition error resilience algorithms. Experimental resultswithH. 263+ baseline tested stream through addictive Gauss white noisewireless channel show that the proposed method outperforms reference method in subjective and objective quality obviously with little increasing of coding overhead and computation.关键词：error resilience coding;video coding;error propagation2354|183|0更新时间：2024-05-08
- 摘要：Under the prerequisite of assuring image quality close to full search algorithm (FS), a fast efficient search algorithm—partial distortion search algorithm forblockmotion estimation based on predictivemotion vector field(PMVPDS) is proposed in order to reduce the computational complexity of existent partial distortion algorithms. PMVPDS takes advantages of the adjustable partial distortion criterion(APDC) proposed in this paper, predictive motion vector field technique and half stop technique to find suited points quickly. Simulation results demonstrate thatPMVPDS provides about 2~7 and 0.6~3 times speedup respectively, and achieves better PSNR performance compared to the normalized partial distortion search algorithm(NPDS) and the progressive partial distortion search algorithm(PPDS). Moreover, the image quality and search speed of proposed PMVPDS algorithm are adjustable.关键词：video compression;motion estimation;block-matching;search algorithm2200|191|0更新时间：2024-05-08
- 摘要：Automatic linguistic indexing of pictures is an important but highly challenging problem for researchers in content-based image retrieval.In this paper,a boosting-based automatic linguistic indexing approach is proposed and a linguistic indexing system called BLIR(Boosting for Linguistic indexing Image Retrieval system) is built.It is assumed that images of same semantic meaning can be represented by a model combined with a group of features.2D-MHMM model is found to be such a template for one special kind of color and texture combination,which corresponds to one cluster in feature space.Thus in BLIR system, a large number of 2D-MHMM models are generated and a boosting algorithm is used to associate keyword with models.The system has been implemented and tested on a photographic image database of about(60 000) images.Results demonstrate the effectiveness of the proposed technique which outperforms other approaches.关键词：content-based image retrieval;linguistic indexing of pictures;Bootsting algorithm;two-dimensional multi-resolution hidden Markov model(2D-MHMM)3173|215|0更新时间：2024-05-08
- 摘要：Kansei image retrieval is a new kind of retrieval technology with high complexity.However,(it's) likely that only some parts of the image would attract people and produce affections.Color imposes a great impact upon the feeling as the basic feature of image,and the difference and comparison of the color would make people produce different kansei.Meanwhile,the entropy of the image also exhibits the information quantity and is a measurement of arousing (people's) kansei.In this paper,we present a method of kansei image retrieval utilizing the color and entropy to extract region of interest(ROI).Back propagation neural network is employed to map the color and entropy of ROI to affective feature space.Finally,we show some experimental results.关键词：kansei image retrieval;region of interest(ROI);entropy;neural network2776|198|0更新时间：2024-05-08
- 摘要：Many traditional indexing methods perform poorly in high dimensional vector space.The Vector Approximation File approach overcomes some of the difficulties of dimensionality curse,but it can't support the quadratic form metric.A novel VA-File approach for quadratic form distance is introduced in this paper.By the SVD of similarity matrix,the quadratic form distance can be converted to the Euclidean distance,and the approximation vector can be obtained. The low-dimensional filter algorithm is also applied during the nearest neighbor search.The vectors are first filtered with the low-dimensional approximate distance measure,and then the candidate results are re-computed with high-dimensional distance measure.The experimental results show that it can save the computational time significantly because only a small set of vectors is computed on the high-dimensional distance measure.关键词：dimensionality curse;quadratic form distance;nearest neighbor search;singular value decomposition(SVD);vector approximation2758|210|0更新时间：2024-05-08
- 摘要：This paper presents a texture segmentation algorithm based on directional Gabor filters when orientation characteristics of textures are taken into account.Incorporating into the human visual characteristics,a design approach of optimal directional Gabor filters is proposed.Each texture can be thought of as containing a narrow range of frequency and orientation components.By filtering an input texture image with Gabor filters tuned to the dominant frequency and orientation component of the textures,it is possible to locate each texture.The magnitude of the channel ouput should be large when the texture exhibits the frequency and orientation characteristics to which the channel's Gabor filter is tuned,vice versa.There are weak intraclass dispersion and strong interclass dispersion in the filtered image,and the issue of texture segmentation is translated into that of traditional image segmentation.Experimental results indicate that the proposed algorithm outperforms conventional approaches in terms of both objective measurements and visual evaluation.关键词：texture segmentation;Gabor filters;human visual characteristics3427|256|0更新时间：2024-05-08
- 摘要：Mean shift is an effective iterative procedure that has been used in edge preserving filtering and image segmentation. It is critical to set bandwidth properly for Iteration. However, traditional fixed bandwidth mean shift is sensitive to input data, and current variable bandwidth estimation methods are not popularly accepted for their high time complexity or inconvenience. Here a new bandwidth estimate method has been proposed, namdy, K-means is employed to estimate bandwidth for each sub-region and its effectiveness is indicated by experiments.2785|211|0更新时间：2024-05-08
- 摘要：This paper applies the Harmonic Analysis of Time Series(HANTS) analysis to study the AVHRR Normalized Difference Vegetation Index(NDVI) time series images of China area.The HANTS algorithm can successfully reduce the cloud effect and at the same time output Fourier components.The Fourier components are used to analyze the variation characteristics of vegetation growing cycle and then study the vegetation cover in regional scale.Analyze the Fourier components of NDVI data after HANTS processing and find that the 0th-order harmonic,or the mean NDVI,indicates overall vegetation cover,and the first and second harmonics of the HANTS analysis concisely summarize the amplitude and phase of annual and biannual values of NDVI data.The amplitude of the 1~(st) harmonic indicates the variability of vegetation cover over the year.The phase of the 1~(st) harmonic summarized the timing of vegetation green-up.The 2~(nd) harmonic indicates the strength and timing of any biannual signal.关键词：time series data;NDVI;Fourier component;HANTS;vegetation cover cycle3808|285|0更新时间：2024-05-08
- 摘要：In order to improve water quality retrievals of remotely sensed image accurately, this paper puts forward a neural network model for water quality retrievals using knowledge and remotely sensed image. The model uses remotely sensed image data and water quality related knowledge as input of BP neural network, then trains neural network, after that water quality is retrieved by the trained neural network. The proposed model is applied to the water quality retrievals of Tai Lake in China. In experiment, knowledge used includes Tai Lake geography information knowledge and classification knowledge of water quality by interpretation of TM image. The result of experiment shows that the developed model has more accuracy than the routine linear regression model and traditional neural network model.关键词：remotely sensed image;water quality retrievals;BP neural network;knowledge4273|220|0更新时间：2024-05-08
- 摘要：Traditional spectral based methods have proven inadequate for the classification of high spatial resolution remotely sensed images, due to the lack of consideration of the spatial features of the data. In order to improve the classification effect of high spatial resolution remotely sensed data, the authorpresented a newmothod called SSMC(Spatial and Spectra lMixed Classifier) driven by integrated spatial and spectral features. In this algorithm,we use twofold pyramids to extract the spatial information of images, one is the background pyramid composed of a setofwindows, such as 64×64、32×32、16×16 and 8×8,which can mimic human perception in identifying objects of different scales; and the other is mallat pyramid, which is used here tomake amultiresolution analysis for every level of the background pyramid. The new algorithm is constructed for the classification of high resolution multispectral images, whose typical cases are IKONOS、Quick Bird etc. And in the experiments, the presented method can achieve better accuracy than the conventional ones, so we can conclude that the SSMC method is active and encouraging.关键词：spectral information;spatial parameter;high spatial resolution;SSMC method3532|190|0更新时间：2024-05-08
- 摘要：The watershed topography is a key factor for extraction of the drainage,and it is also a very important topic in the study of digital watershed and the distributed hydrological models.The paper firstly discusses some problems in hydrological analysis by applying the DEM,and describes three kinds of interpolation models to produce the DEM.Then the author introduces an interpolation method which could improve the accuracy of the DEM by judging the stream's flow direction.Finally the interpolation algorithm is applied to generate the DEM by using the elevation and stream data in the area of southwest China.We analyze the results of the experiment,and find that the DEM created by this method not only keeps its precision but also presents a fine surface runoff character and stream pattern.Moreover,by comparing with the stream lines that extracted from different DEMs,the paper concludes that the DEM created by the hydrological improved algorithm is more useful in dealing with the hydrological analysis.关键词：digital elevation models;hydrology;interpolation;precision2624|189|0更新时间：2024-05-08
- 摘要：Based on deep analysis of the imaging mechanism of OpenGL in the sight of projective geometry, this paper have constructed the relationship between photogrammetry exterior and interior parameters and OpenGL projective matrix and proved their internal consistency. Then the transition equations of two imaging processes are derived, With above relationship and derivation, experiments have been conducted to simulate single static image, stereo image pair and motion orbit and pose in OpenGL. Experimental results prove the correctness of the theory proposed in this paper.关键词：OpenGL;imaging matrix;exterior and interior parameters of photogrammetry;simulate camera's pose;space resection3184|199|0更新时间：2024-05-08
- 摘要：Colorization of grayscale image is in general an active and challenging area of research in image processing.In this paper,current colorization algorithms are summarized and the principle of color propagation is analyzed.Then a colorization algorithm based on Laplace equation is proposed.In order to determine the boundary accurately,a weighted difference scheme is applied.In contrast with Saprio's colorization algorithm which is also a partial differential equation based method,the principle of color propagation in our algorithm is different.Furthermore,our algorithm can achieve more vivid colorized result.关键词：grayscale image;colorization;Laplace equation3685|223|0更新时间：2024-05-08
- 摘要：Erasing grid-line is a key in the Chinese handwriting auto-identify system.After analyzing the advantages and disadvantages of the existed grid-line erasing algorithm we present a grid-line erasing algorithm based on the block adjacency graph(BAG) devised by Pavlidis and the stroke reconstruction algorithm in this paper.Firstly,the scanning graph for Chinese handwriting is changed into the BAG by use of the chain structure,and then the proper Connected Sub-graph according to the BAG is searched.These connected sub-graphs are the grid-lines that should be erased.Finally,the strokes,which left by erasing grid-line,are reconstructed.Because it is easy to get the BAG,searching and erasing grid-lines will be very fast.At the same time,our algorithm can reconstruct the strokes after erasing grid-line,and it not only avoids the disturbing information but also remains the useful characters of Chinese handwriting.Experimental results show our algorithm has strong ability in anti-disturbance and good effect in erasing grid-lines.关键词：block adjacency graph(BAG);erasing grid-lines;connected sub-graph;stroke reconstruction2985|198|0更新时间：2024-05-08
- 摘要：Image mapping and image reconstruction are two major components in performing Image warping.Existing reconstruction methods either involve huge computations or tend to produce visual errors in objects appearance using traditional warping model.This article presents a new image warping model by employing a novel "equal-density mapping" and "texture-Reconstruction".With the new model,a layered image mapping(LIW) approach is proposed then.In LIW,reference pixels are firstly mapped to "equal density layers" that parallel the desired view and are stored in a bidirectional link list after some by-layer processing.By implementing post-texture rendering along the list eventually,the warping is transformed to a fast texture mapping process,which provides a considerable speedup.Also,LIW is proved to efficiently tickle the "rubber sheet" problem without any pre-process.关键词：virtual-reality;image-based rendering;image warping;equal-density mapping;texture-reconstruction;layered2675|504|0更新时间：2024-05-08
- 摘要：Integral imaging is a technique capable of displaying 3D images with continuous parallax in full natural color. Since micro-lens sheet is used in recording, only one recording is necessary in containing 3D information. The paper presents a method for extracting depth information directly from lenticular integral images. The method involves first extracting viewpoint images from the lenticular integral image. The approach starts with the constructions of special viewpoint-images from the integral image. A viewpoint-image is constructed by extracting one pixel from each micro-lens rather than a macro-block of pixels corresponding to a micro-lens unit. Each viewpoint-image contains a 2D parallel recording of the 3D scene from one particular view direction. A new mathematical expression giving the relationship between object depth and the corresponding viewpoint image pair displacement is derived by geometrically analyzing the 3D image recording process. The depth information for any object point can be calculated with its disparity between corresponding viewpoint images. The method is validated and applied to measure the thickness of a matchbox in the experiment, The results can not only be applied on the area of integral image data processing but also be expected to provide theory foundation for a novel depth extraction technique.关键词：depth measurement;3D integral image;machine vision;viewpoint image;disparity analysis3030|225|0更新时间：2024-05-08
- 摘要：A method of inspecting stereo contrast ratio and view zones of autostereoscopic displays using CCD cams is proposed. Firstly, a designed computer graphics is displayed on the screen of autostereoscopic display. Secondly, one CCD cam, instead of the human two eyes, perceives the stereo image pair. Thirdly, the acquired digital images are processed. Finally, the feature parameters extracted from the images are used to calculate the separating degree that indicates the performance of autostereoscopic display. Experimental results demonstrate that the method proposed in this paper can well inspect the stereo contrast ratio and view zones of autostereoscopic display.关键词：autostereoscopic display;virtual reality;stereo contrast ratio;image quality2778|194|0更新时间：2024-05-08
- 摘要：In this article,6S model was used to calculate the satellite observing reflectivity and radiability above ocean surface with typical oceanic aerosol parameters selected.The result shows that some obvious errors will be induced when visibility drops under 5km.Then a synthetic influence on radiation transmission of aerosol and vapor was considered in the radiation transmission model and equations between whole layer transmissibility and planet reflection was built.Three days of MODIS(moderate resolution imaging spectroradiometer)data corresponding to high visibility,moderate visibility and low visibility were taken for calculation.The result shows close accordance with the observed visibility of the time.This study gives an applicable methodology for transmissibility calculation under different atmosphere condition and provides a potential method for quantitative visibility motoring with remote sensing.关键词：MODIS;radiation transmission model;atmospheric transmissibility;vapor;aerosol3305|210|0更新时间：2024-05-08
- 摘要：A human face recognition technique called modular 2DPCA is presented in this paper.First,the original images are divided into modular images in proposed approach.Then,an image covariance matrix is constructed directly using the sub-images,and its eigenvectors are derived for image feature extraction.Compared with previous techniques based on image vectors such as PCA,there are two advantages for this way: 1)the sub-image matrices don't need to be transformed into vectors prior to feature extraction,and dimension reduction of discriminant features can be effected conveniently;2)singular value decomposition of matrix is absolutely avoided in the process of feature extraction so the features for recognition can be gained easily.Moreover,2DPCA is the special case of modular 2DPCA.To test modular 2DPCA and evaluate its performance,a series of experiments were performed on two human face image databases: ORL and NJUST603 human face databases.The experimental results indicated that the recognition performance of modular 2DPCA is superior to that of PCA and is more robust than that of 2DPCA as well.关键词：linear discriminant analysis;modular two-dimensional principal component analysis(modular 2DPCA);feature extraction;face recognition4556|227|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