最新刊期
卷 12 ， 期 10 ， 2007
- 摘要：In order to solve the coverage problem of digital TV in tunnel, we can adopt the antenna or radiating cable. In this paper we made comparison between the two modes according to theoretical and practical test. We also introduced the advantage and disadvantage based on the two modes, and suitable application of each one.关键词：digital television;tunnel coverage3845|200|0更新时间：2024-05-07
- 摘要：It is a key technique to generate a virtual view in 3D Video conference. In this paper, a new algorithm based on DIBR and image mergence is proposed. Firstly the depth images are smoothed by an edge dependent filter and the reference images are normalized so as to reduce the big holes and the artificial edges in desired image. Then an arbitrary virtual view is generated by making use of 3D Image Warping. In order to deal with the visibility problem quickly, the pixels in reference image are processed in an occlusion-compatible order. Then two generated images are merged to eliminate the big holes in the desired view. Finally the remaining holes are filled with an interpolation method. Experimental results show that the proposed method can obtain satisfactory rendering results.关键词：image based rendering;depth image based rendering;3D image warping;occlusion-compatible order3677|234|0更新时间：2024-05-07
- 摘要：Video streaming applications provide on-demand video file playback services through the computer network. WMV is a common file format in IPTV applications. In order to support WMV formaton Linux orUnix-like platforms novel master-slave mode architecture is proposed in this paper. Based on this architecture, we designed and implement a WMV video streaming system on Linux platforms. Inside the server, we embedded several novel algorithms such as the RTSP query schedule algorithm to improve the performance. The whole system is successfully deployed in the Shanghai’s Archive Library.关键词：video streaming;internet-protocol television(IPTV);windows media videol(WMV);high performance;Linux platform3054|230|0更新时间：2024-05-07
- 摘要：IPTV live streaming systems with client-sever based content delivery infrastructures cannot provide service to mass clients due to their high requirement of server hardware configuration and network bandwidth. This paper presents a P2P(point-to-point) based live streaming system-“ClearLive”, which tries to achieve better scalability than client-server based system. While current popular P2P IPTV systems supportonly PC client, our system mainly supports STB clientand is compatible with PC client.“ClearLive”is properly designed according to the characteristic of STB client and application environment of IPTV. The test result of system shows that“ClearLive”can provide high quality live streaming service to vast number of clients with low cost.关键词：IPTV;P2P;Live Streaming;STB(set-top box)3071|196|0更新时间：2024-05-07
- 摘要：In order to provide high-speed broad band multimedia communications with QoS, an MMWO(Mobile Multimedia communication in Wireless Optical networks) model is advanced. In this mode,l FSO technology is used, the practical application of multimedia communication with special requirements is fully considered and a quality-of-service(QoS) routing algorithm is provided. Results of simulations show that the algorithm has better performance and adapt to requests for fastmovement ofnodes, bandwidth and high speed in wirelessmultimedia communication.关键词：mobile multimedia communication;wireless optical networks;FSO technology;QoS routing3148|196|0更新时间：2024-05-07
- 摘要：To meet the requirement of multimedia communication, the paper presents a new mobility model(FSOM). FSOM takes special requirements of mobile multimedia services into consideration, and is fit for the characteristics of Free-space Optical Networks. A routing algorithm, using FSOM, is proposed for multimedia services in Free-space Optical Networks. This routing algorithm called TCR is based on topology control algorithm(TTC). TTC controls topology using improved triangulation algorithm. After executing TTC, we acquire Virtual Graph, which increases the average throughput. Formass data and real-time requirement of multimedia communication, TCR combines the proactive and reactive routing, which is based on VirtualGraph.关键词：mobile multimedia communication;Free-space Optical network;topology control;routing algorithm2736|183|0更新时间：2024-05-07
- 摘要：In this paper, we applied ALM(application layermulticast) into HDTV live system, design and implement a multimedia live prototype system with large concurrency, high efficiency and low delay. We also created a distributed time-shifting algorithm, which was integrated in the system in order to allow users to watch both live and delayed TV programs. This prototype system was put into large-scale testing formore than one month in Fudan Campus during World Cup 2006. The result shows that it is a good solution to output bandwidth bottleneck at server side and long delay of normal P2P technology at the client. The maximum amount of online users was over 460. It also provides a good solution to HDTV live system, which is the focus of National Science and Technology Department.关键词：ALM;multimedia live system;distributed time-shifting2627|206|0更新时间：2024-05-07
- 摘要：Under the wireless network, the performance of TCP is better than that of UDP which lack of congestion control. In order to conform the RTP-over-TCP transmission mode to our hybrid centralized-distributed server architecture, we propose a method with double-TCP channel. Besides, we compare the transmit performance of TCP blocking mode to nonblocking mode. Moreover, we try to improve the bandwidth usage of TCP under the wireless network by using muhi-TCP transmission technology. The simulation results demonstrate that our method is applicable.3312|224|0更新时间：2024-05-07
- 摘要：A uniform 9/7 and 5/3 DWT filter by lifting scheme is proposed for JPEG2000. This paper also proposes a control method of row and column parallel filtering with minimum five lines cache. This architecture can achieve higher hardware utilization and lower temporary data storage without losing precision. By using this architecture, we have accomplished the hardware design. The algorithm is described by Verilog HDL, and simulated by modelsim, It is implemented by cyclone2-EP2C35FC672 under Altera DE2 board with Quartus 6. 0.3058|207|0更新时间：2024-05-07
- 摘要：To adapt the intra predictor in multi-standard video coder to H. 264 and AVS stand, a configurable hardware architecture design of intra predictor that supports both H. 264 and AVS was proposed based on the similarities bebween the intra prediction algorithms in both standards. This kind of architecture can implement most intra prediction algorithms in one configurable calculate unit and thus can reduce the use of chip area. In order to enhance the speed of process, four parallel configurable generator unit are adopted in our design. This design was synthesized and the result showed that this architecture occupied 10 371 LUTs on FPGA and its frequency can be 150 MHz.关键词：H. 264;AVS;intra prediction;hardware architecture;configurable2812|210|0更新时间：2024-05-07
- 摘要：In this paper, a hardware architecture forquarter-pixel interpolation in H.264 is proposed. As for the luminance component, we adopta design which is using four parellel filters which can turn 2-D filter into 1-D filter and using pipeline technique and shift add instead of multiplication in the mean time. While for the chrominance component, we use a design comprising of only shift add to calculate the eight-pixel interpolation. The architecture proposed in this paper can meet the requirement of SD and HD and it works fast and consumes less resources.关键词：quarter-pixel interpolation;hardware architecture;filter;shift and add3070|214|0更新时间：2024-05-07
- 摘要：In the video-encoding process, the encoding time is greatly influenced by many factors, such as inter/intra mode decision, motion estimation(ME), etc. In order to encode the videowith faster speed and beteer quality, this paper only focuses on the intra mode decision in H.264 and presents a fast intra prediction mode decision algorithm for 4×4 blocks in H.264. The proposed algorithm uses: (i) high correlation of RD cost between best prediction mode and its neighborhood prediction mode, (ii) strong correlation between sums of absolute transform differences(SATD) and rate-distortion(RD) performance, efficiently skipping some unlikely prediction modes. Thus, it can complete 4×4 intra prediction by only computing RD cost for four times. The simulation result shows that the proposed algorithm is a good compromise between the coding performance and the encoding speed.关键词：H. 264;intra prediction;video coding2942|224|0更新时间：2024-05-07
- 摘要：A new fast and adaptive intermode decision algorithm for Pslices in H.264 is proposed in this paper. In the algorithm, SKIP mode is decided to be the best mode without Lagrange rate-distortion optimization(RDO), P8×8 mode is decided to be removed from the candidate modes before Lagrange RDO. With the above decision of the two special modes, the entire encoding time is greatly reduced. Simulation results show that the proposed algorithm can save 60%~80% entire encoding time with negligible variance in PSNR and bit rate compared to the H.264 reference software JM 10.2.关键词：mode decision;H. 264;adaptive3038|205|0更新时间：2024-05-07
- 摘要：Considering user access preference in streaming media, the paper proposes a new proxy caching replacement algorithm based on prefix caching and media popularity. The algorithm calculated the predicted popularity using the external and internal popularity of media segment. Then the segment with low popularity is chosen to be replaced. Simulation results demonstrate that, the method can achieve high caching hit rate and reduce request delay under same conditions.关键词：streaming media;caching proxy;popularity;prefix caching3267|220|0更新时间：2024-05-07
- 摘要：In this paper, a robust content-based color image retrieval algorithm using multiple features is proposed. This algorithm takes into consideration the important image edges and utilizes the support vector regression(SVR) theory. Firstly, the image denoising and color edge detection are performed using the SVR classification. Secondly, the whole color edge is divided into local grids; and the color histograms and texture histograms for local grids are computed as image features. Finally, the similarity between color images is computed by using a combined feature index based on the color histogram and texture histogram for local grids. Experimental experiments, including comparisons with many state-of-the-arts, show the effectiveness of our algorithm in improving the retrieval performance(especially for noisy images).关键词：image retrieval;support vector regression(SVR);color edge;noise3027|218|0更新时间：2024-05-07
- 摘要：With the development of motion capture techniques, more 3D motion libraries become available. In this paper, space transformation property and time property are extracted from human motion capture data, which compose 3D motion time sequence features. Given the assumption of the features of each joint is independent, a revised data-driven decision tree of anytime induction is automatically constructed to present the influence of each point during the comparison of motion similarity. Experiment results show that our approaches are effective for motion data retrieval.关键词：motion capture;3D time sequence feature;decision tree;retrieval2816|226|0更新时间：2024-05-07
- 摘要：The convenient attribute of cameraphone makes it popular nowadays: people can use it to snap picture at any time and everywhere. At the same time with the advantage of inherent network connectivity, we can share the snapped image with others immediately or search for similar images from the web. The above advantages of cameraphone make the quantity of cameraphone images boosting quickly, which results in the problem of how to efficient organize, manage and retrieve. In this paper, a new region based cameraphone images retrieval incorporated with metadata method is proposed to improve the image retrieval performance. Using this algorithm, we implemented an online demo system to allow users to upload their snapped images and do searching similar images. Compared to traditional contented based image retrieval (CBIR) method, ourmethod uses the region feature of the image incorporated withmetadata of cameraphone to bridge the semantic gap between the low-level feature of images and semantics. The experimental results show higher efficiency ofour methods than the traditional CBIR method.关键词：image retrieval;contented based image retrieval(CBIR);region basedmethod;cameraphone;Exif2759|225|0更新时间：2024-05-07
- 摘要：In this paper we firstly present an XML-driven image semantic retrieval system, which consists of several functional modules, i.e. image content hierarchy modeling, semantic object automatic extraction, and semantic similarity measure. Secondly, the paper concentrates on the image character description model, and we discuss the definitions and operators of image object. Then some related XML semantic descriptors are defined to represent the image content. Experiment results on traffic images show good retrieval quality.关键词：semantic retrieval;XML descriptor;content hierarchy;image retfievall3264|197|0更新时间：2024-05-07
- 摘要：In this paper, a geometrically robust image watermarking algorithm based on significant region of normalized image is proposed. The geometrically invariant space is firstly constructed by using image normalization method. Then the significant region is obtained from the normalized image by utilizing the invariant centroid theory. The digital watermark is embedded into the significant region by quantizing the magnitudes of the selected wavelet coefficients. Experimental results show that the proposed scheme is not only invisible and robust against common signals processing, but also robust against the geometric distortion.关键词：image watermarking;geometric distortion;image normalization;significant region2975|205|0更新时间：2024-05-07
- 摘要：In this paper a new unequal error protection method for IPTV transport stream is introduced. We also proposed a new evaluation standard (ENCS), and based on ENCS, a new interlacing strategy. It efficiently improved the transportation reliability over packet-loss network, and reduced the impact from packet loss on video quality.关键词：unequal error protection(UEP);transport stream(TS);data interlace;IPTV;PES;PSI2688|169|0更新时间：2024-05-07
- 摘要：The error concealment technique can efficiently compensate the data loss and the data errors in the transmission process caused by the transmission environment, thus improve the quality of the video received. In this paper, we propose a novel adaptive temporal error concealment technique~ To conceal a lost block, we estimate both the motion of the whole block and the motion of every pixel in the block, and then adaptively combine the two estimated motion to obtain the recovery value of the missing pixel according to the status of a missing pixel in this block. Experimental results show the proposed method is significantly better than the existing techniques in terms of both subjective and objective quality.关键词：video transmission;error concealment2868|206|0更新时间：2024-05-07
- 摘要：Autostereoscopic technology and shutter glasses have been applied inmany areas. 3D images can be perceived based on the principle ofHVS(human visual system), which is also the foundation of visual cryptography. However, the traditional method of visual cryptography cannot achieve satisfactory result because of new features of 3D video technology. In this paperwe proposes a new method of visual cryptography combined with the characteristic of 3D video display,which takes the advantage of disparity in HVS to hide the secret information in right image. Simulation results confirm the validity of the proposed method.关键词：3D display;visual cryptography;autostereoscopic techonology;shutter glasses2845|206|0更新时间：2024-05-07
- 摘要：This paper exploits the distribution of energy and correlation between various components of sub-bands and modifyies the SPIHT algorithm. In this scheme, insignificant coefficients of first component in high-frequency sub-bands that correspond to the same spatial location in other components can be used to reduce the scanning scope in the process of invalid set discrimination. Both theoretic analysis and experiment results confirm good compression performance and low complexity at the same time.关键词：image compression;SPIHT algorithm;coefficients rearrangement;multiwavelets transform3064|168|0更新时间：2024-05-07
- 摘要：AVS1-P2 introduces several prediction modes for inter coding. In order to achieve the best coding efficiency, the rate distortion optimization is exploited to choose the most suitable predicting mode, but it is very computationally expensive. In this paper, according to the characteristics of AVS1-P2 integer transform, we propose a new fast inter mode decision algorithm to improve the encoder speed based on two criterions, early termination based on all zero block detection and the prediction of inter-mode size using texture characteristic of macroblock. Compared with the reference code, simulation results show that this method will reduce the computing complexity of encoder and improve the encoding efficiency of AVS1-P2 software encoder while bit-rate and video quality almost remain unchanged.3476|220|0更新时间：2024-05-07
- 摘要：JM is the standard testmodel published by joint video team(JVT), which has great significance in algorithm learning and research. In JM, there are three optionalmodes by setting parameters differently. Based on the study of the principle of the UMHexagonS algorithm, an improved method is proposed to reduce the time ofmotion estimation with minimal loss in bitrate and reconstructed quality compared with those of the originalUMHexagonS algorithm. Correlation between adjacent blocks is used to end the search. In JM10.2, results show that the improved UMHexagonS algorithm can reduce the time of motion estimation with the original performance of the original UMHexagonS algorithm.关键词：H. 264;motion estimation;UMHexagonS algorithm;JM test mode3441|188|0更新时间：2024-05-07
- 摘要：This paper presents an error protection technique for SPIHT-coded image transmission in etTor-prone network. For the lowest and next lowest frequency coefficients, protection scheme based data hiding is performed. The lowest and next lowest frequency coefficients, which are taken as the hidden data, are extracted from the compressed bitstream, and embedded back into the same bitstream. The restored hidden data is used to conceal errors. The coefficients that cannot be recovered in other high frequency coefficients reconstruction are predicted through linear interpolation based on intrasubband or inter-subband correlation. Experimental results show that the proposed method achieves good and stable error performance with minimal additional redundancy.关键词：SPIHT-eoded;true 2D wavelet;error protection;data hiding;error concealment3631|201|0更新时间：2024-05-07
- 摘要：To lower the complexity of bit plane scanning arithmetic coding in image compression, arithmetic coding can be isolated from bit plane scanning for the purpose of parallel computing. The MQ-coding in JPEG2000 is isolated from the bit plane scanning. Based on the characteristic of coding algorithm after isolated coding, a modified MQ-coder is designed. Also, this paper studies the implementation in TMS320C6000 DSP of the modified algorithm. The processing speed is improved through software pipelining, loop unrolling, conditional instruction and complier options. The coding performance of arithmetic coding is superior and its speed is improved by eight times due to the implementation of isolation coding and modified MQ-coder.关键词：image compression;arithmetic coding;parallel computing;isolation coding;modified MQ-coder4399|216|0更新时间：2024-05-07
- 摘要：There is a great demand to support high resolution local video conference in today’s PCs. However, although PC is convenient and local access network bandwidth is sufficient, the computational capacity is a big bottleneck in high resolution laptop video conference system. Here, Template-based Coding Algorithm(TBCA) is proposed for such a situation which can mi prove the utilization of computing resources by 10% and reduce the adjustment frequency to half of the initial situation. A high resolution laptop video conference system based on the TBCA is miplemented with satisfactory user experiences.关键词：adaptive coding;live video system;high resolution;video coding3725|193|0更新时间：2024-05-07
- 摘要：With the development of video encoder and decoder technology, H. 264 has become one of the most important standards. In this paper, we proposed novel moving object detection based on spatial-temporal method in compressed domain, which mainly depended on the motion vector field and efficiently segmented the object. Firstly, a spatially median filter was applied to the motion vector to alleviate the error vector due to difference between the motion vector obtained from motion estimation and the real motion information. Secondly, we built the spatial and temporal coherence model with the entropy based method. We obtain the threshold with these models. The effectiveness of our method was demonstrated with its performance on experiment results.关键词：compressed domain;object detection;H. 2643367|196|0更新时间：2024-05-07
- 摘要：Moving object segmentation in the compressed domain is fast and efficient compared to former segmentation in pixel domain. Recently some moving object segmentation algorithm in MPEG compressed domain have been reported. However, very few approaches have been proposed in the H. 264 compressed domain. In this paper, a real-time video segmentation method in H.264 compressed domain is proposed. Firstly, the current frame is globally compensated in motion after the globalmotion estimation. Then the 4×4 motion vectors are processed bymedian filter and classification. Thirdly the non-zeromotion vectors are applied by improved EM clustering segmentation algorithm. The performance of the proposed approach is illustrated by simulation carried on several test sequences. Experiment results show that the proposed algorithm can provide satisfactory real-time segmentation results both on static background and moving background video.关键词：H.264;compressed domain;moving object segmentation;EM clustering;global motion compensation2974|243|0更新时间：2024-05-07
- 摘要：A new multi-description video coding scheme based on distributed source coding is proposed to solve the drift problem of the multi-description video codec under packet loss channels. Experimental results demonstrate that the proposed scheme can effectively avoid the drift problem by virtue of the robustness of distributed source coding.关键词：distributed source coding;multi-description video coding;packet loss channel;drift problem3086|217|0更新时间：2024-05-07
- 摘要：The scalable extension of H.264 video coding standard (SVC) has achievesd temporal scalability using Hierarchical-B structure. Compared to traditional coding structures like IPPP and IBBP, ithas a significant coding gain as well. However, due to its hierarchical prediction dependency, quality fluctuation among consecutive frames may sometimes occur. Such situation usually gets worse when high motion or scene changes exist. In this paper, we propose a simple but effective constant quality algorithm based on PSNR and MAD measures. Restriction to the possible fluctuation is realized by adapting the quantization parameter(QP) at frame level. We simulate our method based on JVT official specification. Experimental results show that our algorithm has greatly reduced the PSNR variance and obtained the target PSNR simultaneously.关键词：constant quality;Hierarchical-B structure;scalable video coding2938|194|0更新时间：2024-05-07
- 摘要：When capturing panoramic scene, the camera usually moves only in the horizontal direction. If we encode the captured panoramic video directly, the encoding efficiency will be low dueto the possible light dithering of the camera in the vertical direction and noise introduced by sensor. This paper presents a highly efficient algorithm for encoding panoramic video with H. 264 standard according to its special characteristics. First, we convert the original panoramic video to a panoramic image, Then we can convert the panoramic image to H. 264 bit-stream directly. Compared to the encoding of the original panoramic video, the algorithm we present highly improve the encoding efficiency.3954|186|0更新时间：2024-05-07
- 摘要：To get rid of the interference of shadows when detecting objects, we discussed the theory of illuminant invariance image and its important property first. Based on this theory, a method for removing the casting shadows of objects was proposed. According to optical imaging theory, this method transforms a color image to an illuminant invariance gray- scale image and removes the shadows. To obtain the illuminant invariance image, a camera-dependent illuminant invariance angle should be measured first. Aiming at solving this inconvenient, a histogram-based rule for deciding the illuminant invariance angle was proposed also. A number of experiments results upon different scenes demonstrated that, the proposed method is able to remove the casting shadows effectively and segment the objects accurately.关键词：illuminant invariance image;illuminant invariance angle;histogram statistic;shadow removal4046|240|0更新时间：2024-05-07
- 摘要：In this paper, we introduce an effective algorithm based on blocks (BoB) for eye location. The image is partitioned to blocks by a threshold. Based on the comparability between two eyes and the uniqueness of a pair of eyes, it can locate eyes after matching each two blocks. Due to what we use are the location and the comparability between two eyes and eyes are considered in pairs, the algorithm could work on rotated images and get higher detection rate. What’s more, it can deal with moderate illumination and expression changes. Experiments on a number of face images have shown that our algorithm is reliable and fast.关键词：facial feature;eye localization;face recognition;block3126|235|0更新时间：2024-05-07
- 摘要：Developing a reliable face recognition system has been an open problem. Infrared (IR) imagery appears to be a good alternative as it is independent of external illumination. However, IR images are also affected by a series of factors ranging from imaging, ambient condition to physiological and psychological conditions. The variations of temperature in IR images have analogous effects of illumination in visual images. In this paper, a method based on Stefan-Boltzmann Law has been applied to convert the facial temperature data intomore consistent radiation energy data. The experiments conducted illustrate that the radiation energy data are less sensitive to ambient temperature if the human bodies are in steady state, and significantly out performs temperature data in terms of recognition rate.关键词：face recognition;infrared images;radiation energy3499|213|0更新时间：2024-05-07
- 摘要：It is difficult to obtain the accurate segmentation threshold for image. This paper uses the trigonometric series to approximate the histogram and analyzes the fitted curve of the histogram. The problem of the optimal threshold has been solved by estimating the parameters of the inflexion of the curve which approximated the histogram. The experimental result shows that this method is effective in calculating the optimal segmentation threshold of an image.关键词：trigonometric series;histogram approximation;function inflexion;optimal threshold3224|214|0更新时间：2024-05-07
- 摘要：From the point of word image recognition application in a cell phone. We analyzed the tunnel-effect feature which popularly existed in a cell phone and proposed a method which can abstract image from the word-contained images. The established model can rejuvenate the input image so as to finally resist vignetting from degradation of camera pattern in a cell phone.关键词：character recognize;smart phone;binarization;regression model;dark-corner correcting3204|212|0更新时间：2024-05-07
- 摘要：Image fusion technology should try to achieve two objectives: one is to combine useful information of source images asmuch as possible; the other is to abandon the useless information of source images. Based on the analysis of blind signal separation theory, an ICA-based image fusion algorithm is proposed in this paper. Firstly, some pre-processing methods and ICA analysis are applied to each source image; then the ICA coefficients computed in the first step are fused using the maximum rule. Finally the fused image is obtained by synthesizing the fused ICA coefficients. Experiments prove that the new algorithm is feasible and effective formulti-source images fusion.关键词：multi-source image;image fusion;independent component analysis(ICA);matrix sparsity3973|189|0更新时间：2024-05-07
- 摘要：An auto-compositor arithmetic of image sequence based on phase correlation is presented in this paper. We use phase correlation to estimate if two images overlapped, and ascertain the relationship of the overlapped images. This algorithm avoids intervening in image sequence by man. The experimental results have shown that the algorithm is effectively, and can be used in Image Mosaics well.关键词：panorama image;image mosaic;image sequence;auto-compositor;phase correlation3332|239|0更新时间：2024-05-07
- 摘要：Choosing a good Interpolating template is the key to better recover the subsampled image with low distortion. The All Phase Digital Filter is a new type of linear phase filter. According to All Phase theory, we propose 2-D Walsh interpolating template. Here a newmethod is applied: first diagonal interpolation, second horizontal interpolation and vertical interpolation. This algorithm has been compared with the nearest interpolation, bilinear interpolation, spline interpolation. Simulation results show this algorithm is with less computation complexity and can reduce the influences from noise.关键词：all phase digital filter;interpolation;super resolution3095|178|0更新时间：2024-05-07
- 摘要：A fast and robustApproach for the detection of text on road signs based on improved SCW (sliding concentric windows) is proposed in this paper. The algorithm mainly contains three stages. We first convert the input image to a binary image using improved SCW then we obtain the blob information using CCA (connected componentanalysis), and at last we achieve the final text location on road signs by a number of judging criterions. Experimental results demonstrate the great robustness and efficiency of our method.关键词：SCW;CCA;character location;vehicle navigation2892|208|0更新时间：2024-05-07
- 摘要：W ith the application o fData Matrix in people’s daily life and industrial product, the detection of Data Matrix has become very useful. The existing detection algorithm is only suitable for printed paper; it can notbe used to detect Data Matrix which is punched on other material surfaces. This paperpresents a machine learning based on the method which can detect the Data Matrix on various surfaces. We extend the Ada Boost algorithm to Adaptive-Spatial Boost which can adaptively combine the texture information and spatial connection. The Data Matrix detector trained by this algorithm has achieved 100% detection rate on the test samples in our experiment.关键词：data Matrix;machine learning;Adaptive SpatialBoost;multi resolution3022|223|0更新时间：2024-05-07
- 摘要：In order to solve the problem that the low-level visual features can not express the high-level semantic contents accurately in image retrieval, the paper presents a new image similarity calculating approach using the proposed semantic measure. Based on the segmented image regions, the statistical mapping relationships between image blobs and semantic metadata are constructed and the layered semantic measurements between images are then defined. Experiment results on natural images show good retrieval quality based on the semantic similarity measure method.关键词：semantic measure;image similarity;image semantic;image retrieval3223|203|0更新时间：2024-05-07
- 摘要：Accurate texture mapping is a key factor for reconstructing visually realistic models. After basic theory of texture mapping is introduced, two practical issues are discussed for 3D reconstruction based on 3D laser scanner. An algorithm of model reconstruction based on its texture and an algorithm of lighting consistency recovery of texture images are proposed to deal with the inconsistency of images in space and lighting. Experiments with real outdoor data prove the effectiveness of these algorithms, which will lay a technological foundation for improving visual reality of 3D reconstruction of outdoor objects in large scale.关键词：reconstruction;texture mapping;lighting consistency4351|223|0更新时间：2024-05-07
- 摘要：To solve the complex and nonlinear illumination direction classification problem in face recognition, a method based on bifurcate tree and SVM is proposed in this paper. Before using the SVM, light directions are firstly sorted into a bifurcate tree according to its intrinsic intensity distribution which assure that only a few classes are sorted at every branch. Moreover, different features are extracted for SVM at the different levels according to the lighting characteristics of the corresponding levels. Experiment results show that the proposed method is efficient to solve multi-classes classification problem and the accuracy can be achieved to 89.16%.3493|228|0更新时间：2024-05-07
- 摘要：Multifractal spectrum algorithm based on multi-weight method(MWM) was presented in a accordance with the presnt multifractal spectrum sensitive to the noises. The definition of multifractal spectrum based onmeasurement theorywas introduced. The probability measure based on MWM was defined that can be used to estimate the multifractal spectrum. The capability of multifractal spectrum of MWM was analyzed. In the paper we analyzed the linear region, the anti-noise performance and the convergence of MWM compared to the Bottom interface and variance. The above approach showed that the noise can be reduced effectively and the fine property can be stressed. The local property of an image can remained muchmore and the weighted factor is converged quickly than the approach based on Bottom interface and variance.关键词：fractal;multifractal;edge detection;image segmentation2891|214|0更新时间：2024-05-07
- 摘要：The fingerprint segmentation is aimed to exclude non-ridge regions and unrecoverable ridge regions and remaining ridges as background. This could improve the fingerprint recognition accuracy and image processing speed. In general, remaining ridges are of clear structure. Which makes it difficult to identify remaining ridges from the interested regions merely by characteristic parameters of fingerprint image. The paper proposes two steps for fingerprint segmentation aiming at removing remaining ridges from background. If the remaining and the real ridge regions are separate to each other, the area parameter is used to remove remaining ridges in the first step, In the case of conglutination, the shape features of the fingerprint image are aiming to find the optimal segmentation point in the second step. The experimental results show that the proposed method leads to adequate accuracy of minutiae detection.关键词：fingerprint;image segmentation;remaining ridges;line label;chain code3236|228|0更新时间：2024-05-07
- 摘要：The paper discusses the application of Principle Component Analysis(PCA) in image’s feature attributes reduction. After PCA pre-processing, Rough Set theory was introduced, and its application in characterized parameters’ attribute optimization was also explored. The unnecessary attributeswere eliminated with an attribute reduction algorithm. The inner redundancy of CBIR was revealed. The result of attribute reduction using UCI dataset proved the algorithm can exclude the influence of unused attributes and decrease the complexity of CBIR effectively.关键词：PCA;image;rough sets;reduction2858|199|0更新时间：2024-05-07
- 摘要：An optimized design method of the infrared image processing system is proposed in this paper. The system adopts INFRARED-2500AS as the infrared detector. Based on a high performance digital media processor DM642, the captured infrared images are processed with several algorithms, such as image interpolation and histogram equalization. Video encoder is used in order to output infrared images. After explaining the system structure and the instruction process flow, the corresponding optimizing strategies of the hardware and the software is introduced, including system circuit design and an improved memory structure allocation. The multimedia instruction set of DM642 is also used for code optimization. Experimental results show that the speed and stability of the system have been remarkably improved, which could meet the requirements of large data volume, high-speed transmission and complex operations of the infrared image processing algorithm. The system can be applied in poor environments and will play an important role in the military and civilian fields, etc.关键词：DM642;infrared image;optimization;high-speed circuit;memory allocation2909|220|0更新时间：2024-05-07
- 摘要：Pedestrian detection is a difficult problem in the field of object detection. This paper applies the boosted cascade algorithm which originates from face detection to pedestrian detection. We improve the algorithm by first project the high dimensional hog(histogram of gradient) feature to 1 dimension, then estimate the posterior probability with lookup table so as to apply the gentle adaboost algorithm. Experiments indicate our algorithm is both efficient in training and detection. The detection rate is close to the state of the arts.3823|217|0更新时间：2024-05-07
- 摘要：A primary class of the image inpainting technique is to build up a partial differential equation, and solve it with some iterative method. One of the most representative inpainting algorithms is BSCB mode.l Because BSCB model’s main drawback is slow calculation, this paper proposes selective adaptive interpolation algorithm based on diffusivity function. The experimental results show that this method is easier to implement and faster in computation than BSCB model, and improves the inpainting effects.关键词：image inpainting;BSCB model;diffusivity function4543|241|0更新时间：2024-05-07
- 摘要：In active vision system, intelligent analysis of objects-of-interest(OOI) in 3D scene generally needs multi-agent cooperation, which depends heavily on the accuracy of multi-view geometry computation and the robust correspondence of OOI. In most cases, active vision system captures multiple views under wide-baseline stereo model, which results in distinct affine distortions between views and complex description of correspondence problem. Furthermore, the difficulty of depicting correspondence problem in active vision is aggravated when the cameras perform real-time pose adjustment for active OOI tracking with the best viewpoint. For above mentioned issues, this paperproposes amethod to achieve real-time multi-view geometry updating on the basis of constructing geometric feature invariants. Then a robust OOI correspondence algorithm is advanced with the acquired multi-view geometry constraint. The experimental results demonstrate that our method can obtain robust OOI correspondence under wide baseline stereomodel and be suitable for real-time application in active vision system.3127|253|0更新时间：2024-05-07
- 摘要：Color quantization or color reduction is an important technique for image analysis and has been widely used in image segmentation, image compression and image recognition. Firstly, on bisecting K-means is used to quantize image roughly and then we refine the image by improved spectral clustering based weighted distance. The stability and quickness of bisecting K-means and adjustable weight make our approach an attractive one. Experimental results show that our approach performs better than octree algorithm in quantized quality and has a less computation complexity than K-means algorithm. For special image, which includes one important color but with only a few pixels, traditional approaches usually lose the important color, but our approach can deal with it by introducing the weight for distances between pixels.关键词：color quantization;spectral clustering;bisecting K-means3207|186|0更新时间：2024-05-07
- 摘要：Self-adaptive Smooth Correct or model(SASC) is proposed for the restoration of the blotch. An algorithm, named bidirectional matching, is designed for motion estimation in the restoration specially. Without making judgment, this algorithm can detect the suitable motion direction on different situations of noise occupation. The Weber ratio is used to make the noise detector to be sensitive and accurate and the bulge algorithm is used to dealwith the edge-problem of the noise. Experimental results using video sequence with real film have demonstrated the availability of the method.关键词：blotch noise;self-adapting detect;Weber ratio;three step search3052|222|0更新时间：2024-05-07
- 关键词：segmentation;region growing;watershed;outlier2786|237|0更新时间：2024-05-07
- 关键词：image;radial distortion;oblique distortion;correction3493|223|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