最新刊期
卷 17 ， 期 6 ， 2012
- 摘要：Age information,as an important personal trait,has great potential in safety surveillance,human-computer interaction,multimedia applications,and face recognition.As an emerging biometric information identification technology,face-image based age estimation has gained great attention resently and has become one of the important research topics in machine learning and computer vision.In this paper,we survey most existing commonly used methods in face-image based age estimation,especially focusing on the extraction of age features and classification.Then,we also introduce some face aging databases and evaluation protocols,which are widely used at present.Based on these databases and evaluation methods,a comparison of the performances of several age estimation systems is presented.Finally,the challenges and promising directions of age estimation techniques are discussed.关键词：face aging;age estimation;age feature extraction;age classification;aging database5041|300|0更新时间：2024-05-08
- 摘要：Expectation maximization (EM)algorithm for parameter estimation of image statistical model is one of the striking research fields in recent decades.Based on the analysis of the EM algorithm,combining the current application research in parameter estimation of image statistical model,analysis and comparison are conducted in terms of the three improvement schemes of standard EM algorithm.In this paper,integrating image restoration,segmentation,object tracking and the fusion of other evolution optimization algorithms,through three aspects,such as the selection of missing data sets,the statistical model establishments of missing and incomplete data sets,and parameter estimation of image statistical models,as well as the advantages and disadvantages of the corresponding EM algorithm are exponded.The structure and complexity of EM algorithm,so far as to success or failure,are directly determined by the selection of missing data and the expression form of incomplete data.In the end,challenges and possible trends are discussed,and extensive applications of EM algorithm to parameter estimation of statistical model with missing data are pointed out.关键词：expectation maximization algorithm;image statistical model;parameter estimation;evolution algorithm3575|151|0更新时间：2024-05-08
- 摘要：Gradient-driven PDEs (partial differential equations) are the main computing pattern for geometric inpainting models of digital images.Apparently,compared with previous models,gradient-driven computing models have a great advantage to the large-scale regions geometric inpainting,but its performances are not stable to different inpainted objects because the information propagating direction is uncertain in the inpainting process.Based on analyzing the computing essences and the corresponding physical meanings of gradient-driven models,it is decisive to the inpainting result that the information propagating direction always points to the outside of the inpainted regions.Thus,a new method of gradient-driven image inpainting is proposed.Experimental results prove that the method can stabilize the information propogating direction making its inpainting performance is more robust.关键词：image restoration;partial differential equations;gradient driving;information propagating direction3247|155|0更新时间：2024-05-08
- 摘要：Compressed images may have block artifacts at low bit rates in many image compression algorithms.Post-processing methods for image de-blocking are the most practical solution for removing block artifacts,since this does not require any changes to the existing standard codecs.Image de-blocking can be considered as recovering the ground-truth image from inaccurate samples.It is exactly what compressive sensing does.According to this,we take advantage of compressive sensing theory to remove block artifacts.As a result,we convert the image de-blocking problem to a linear programming problem in which no parameters are required to be tuned.Finally,our approach can be performed fast using a GPU implementation.Our experiments show our approach can effectively remove block artifacts from compressed images,improving the visual quality and PSNR.关键词：block artifact reduction;compressive sensing;graphics processing unit (GPU);curvelet transform;linear programming2943|103|0更新时间：2024-05-08
- 摘要：A perfectly blind robust quantization-based watermarking scheme in DWT-SVD (discrete wavelet transformation-singular value decomposition)domain is proposed by introducing the self-embedding technology.It can accomplish copyright authentication only resorting to the attacked image. The original image is conducted with DWT, and its low frequency wavelet band is split into non-overlapping blocks. Afterwards, each block is conducted with SVD. A feature watermark sequence is derived through judging the numerical relationship between the two biggest singular values of adjacent blocks. Moreover, the chosen feature watermark sequence is self-embedded into each block’s biggest singular value from the original image’s low frequency wavelet band based on the Principle of odd-even quantization. Finally, a watermarked image is obtained after SVD synthesis and IDWT. The proposed algorithm has good invisibility and security, and achieves perfectly blind detection by combining self-embedding feature watermark sequence and blindly extracting authentication watermark sequence. Experimental results show that the proposed algorithm has strong robustness towards adding Gaussian noise, adding salt and pepper noise, Gaussian low-pass filtering, median filtering, cropping, JPEG compression,and hybrid attacks.关键词：digital watermarking;quantization-based watermarking;perfectly blind detection;self-embedding technology;robustness3329|99|0更新时间：2024-05-08
- 摘要：In order to orientate the function area in brain,an accurate skull extraction method is needed in fMRI image processing.But,we often miss the weak edge information in the skull fMRI image because of the existing imitations of MRI and conditions.For this problem,an efficient weak edge extraction method is proposed in this paper.On the premise of none of the target information losing,we first use mean-shift clustering algorithm to weaken the noise of the image.Then,based on the distribution features of the filtered images,a histogram enhancement method is used to enhance the skull area.At last,use canny edge detection algorithm to detect the edge of the skull in fMRI image.The experiment results represent that the method can extract the weak edge of the skull in fMRI images efficiently.关键词：weak edge detection;Mean-shift;histogram enhancement;fMRI image3438|216|0更新时间：2024-05-08
- 摘要：The lane detection is a key step in unmanned vehicles and lane departure warning system. For improving the reliability of lane detection under complicated situation, such as shadows, damaged pavement and vehicle occlusion, the characters of structure lane that the gradient of both side lane marker have opposite directions is utilized, and the lane detection problem is converted to detect the middle line and width of lane based on gradient-pair constraint. Then the parallel perspective model and linear-hyperbola model are used respectively to estimate the vanishing point coordinate, lane width and parameters of middle line by Hough transform. Compared with the other two algorithms under complicated lane situation in extensive experiments, the results demonstrate the effectiveness of the proposed method.关键词：lane detection;gradient-pair constraint;linear-hyperbola model;parallel perspective model;Hough transform3525|99|0更新时间：2024-05-08
- 摘要：Detecting object contours from natural images plays an important role in machine vision.However,because of the texture edges existing in natural images,it becomes very hard to implement.Relevant research on orientation selective neurons in the primary visual cortex shows,that a mechanism,called non-classical receptive field,can inhibit texture edges and facilitate isolated edges when the visual system processes natural images.Many biologically motivated models have been proposed for contour detection,but they share a common problem which is that some contour elements will be lost if the value of inhibition level is set to high, while some texture edges will be retained if it is set to low.In order to solve this problem,we present a new model, which combines the information from different inhibition levels.It effectively suppresses texture edges and reduces the possibility of losing contour elements.Experimental results show that in comparison with the traditional algorithms,the new algorithm increases performance about ten percent and is more robust.关键词：contour detection;texture edge;non-classical receptive field;multilevel inhibition3055|108|0更新时间：2024-05-08
- 摘要：A projection method is applied to local image blocks through combinations with Vector Quantization(VQ) to generate the projection vector index histogram, which can efficiently represent the distribution and the spacial information of colors. Furthermore, a block-based local binary pattern(LBP) algorithm is proposed, which can effectively extract the structural model of block primitive, avoid instability,and reduce computations compared to the traditional methods. Finally, an image is partitioned into significant regions and non-significant regions based on its saliency map to give the features which are extracted from them more visual sense. Our proposed method is improving the performance by an average of 6.39% compared to the classical index histogram algorithm.关键词：image retrieval;bitmap feature;projection vector;local binary pattern feature;saliency map2972|153|0更新时间：2024-05-08
- 摘要：In intelligent robot design,the traditional machine learning paradigm is commonly used.However,the traditional methods cause problems in visual tasks such as low learning initiative,lack of adaptability with uncertainty and bad expansibility of knowledge and ability.According to the new research direction called cognitive development learning,a visual novelty driven incremental and autonomous visual learning algorithm is proposed,in which the internal motivation is defined as visual novelty which is calculated by online PCA.The autonomous learning and accumulation of knowledge is implemented in the form of updating PCA subspace,which is guided by internally motivated Q-learning using visual novelty.Equipped with the proposed algorithm,a robot makes the next learning decision by judging the novelty between learned knowledge and what is seen now.Experimental results show that the algorithm has the ability of autonomous exploring and learning,actively guiding the robot to learn new knowledge,acquire knowledge and develop intelligence online and in incremental manner.关键词：cognitive development;internal motivation;visual novelty;online principal component analysis;Q-learning3397|79|0更新时间：2024-05-08
- 摘要：In this paper,a targeting algorithm for cooperative multi-soccer robots based on Hough space is proposed.The robot uses the landmark in the field for self-localization based on triangular relationships.Object localization for the robot is achieved by position and pose transformation from the image coordinate systems to the world coordinate system on a simplified multi-link robot model.ZigBee wireless sensor network is established between multiple robots for wireless communication.Transforming the coordinate points of the multi-robot target localization into Hough space,the optimal estimation can be obtained by least square fitting.Then,the ball tracking is achieved by fusing the improved particle filter.Finally,a simulation experiment is carried out on 21-degree-freedom humanoid soccer robots.Test data showed that the precision of the targeting algorithm for cooperative multi-soccer robots had risen by 48% and the efficiency of target tracking was improved on the premise of meeting the real-time requirement.关键词：humanoid soccer robot;Hough space;target localization;multi-robot cooperation;particle filter3255|96|0更新时间：2024-05-08
- 摘要：To obtain bigger swath width, most high-resolution linear push-broom imaging systems use a many CCD chip configuration instead of a single chip configuration. The CCD chips cannot be fixed linear in the focal plane. The question "how to stitch the images obtained by the CCD chips to a seamless single image?" becomes important in the processing of high-resolution satellite images (HRSI). In this paper, an algorithm of inner filed-of-view (FOV) stitching based on the virtual CCD line is shown. With the derivation of the infection to stitching precision caused by height precision, we found that if the virtual CCD is installed near the real CCD, the average height could be used for the stitching algorithm without a loss of precision; otherwise, precise DEM data will be needed. Data from the ALOS/PRISM sensor has been used as test data to verify the algorithm. Through the stitched forward, nadir, and backward images, no position mistakes can be found. A rigorous geometry model of the stitching images also has been made. Comparing the forward intersection results and the original images rigorous sensor model and the stitching images rigorous geometry model, the stitching images has the same orientation accuracy. The ALOS/PRISM images can be stitched without DEM by this method, and the precision of stereoscopic mapping is not being reduced. This method also can be used for combining images from digital aerial cameras.关键词：virtual CCD line;inner FOV stitching;forward Intersection;advanced land observing satellite (ALOS);panchromatic remote-sensing instrument for stereo mapping (PRISM)4624|96|0更新时间：2024-05-08
- 摘要：Digital fingerprinting is a technique for identifying users who use multimedia content for unintended purposes. This paper develops a new fingerprinting scheme based on OFFO (Optimal Focused Fingerprints from Orthogonality)and BIBD (Balanced Incomplete Block Design). This scheme considers OFFO as basic signal and transforms the continuous code into antipodal form of binary code, and uses balanced incomplete block designs to construct multilayer fingerprinting. Theoretical analysis and experimental results demonstrate that, compared with the similar fingerprinting scheme, the proposed scheme can more exactly identify at lest one of colluders with the same user number.关键词：digital fingerprint;optimal focused fingerprints from orthogonality(OFFO);balanced incomplete block designs (BIBD);tracing pirate;collusion attack4953|88|0更新时间：2024-05-08
- 摘要：To extract the discriminant local structural features,an improved infrared face recognition method based on LBP discrimination patterns is proposed in this paper.In traditional uniform patterns of LBP,the most frequency pattern information in nature image is chosen for image recognition.However,the most frequency patterns are not most suitable for face recognition.Based on supervised learn idea,pattern selection algorithm is proposed to get the LBP patterns which are most suitable for infrared face recognition.To make full use of the space locations information,the partitioning and LBP histogram are applied to get final features.The experimental results demonstrate the infrared face recognition method based on LBP and discrimination patterns proposed outperforms the traditional methods based on LBP or PCA.关键词：local binary pattern(LBP);infrared face recognition;pattern selection(PS);separability discriminant (SD);discrimination patterns4764|110|0更新时间：2024-05-08
- 摘要：We propose a new method based on local space-time interest points and self-organization feature maps(SOFM)to recognize and retrieval complex events in real movie.In this method,an individual video sequence is represented as a SOFM density map,We integrate this density map with a support vector machine(SVM)to recognize events.Local space-time features are introduced to capture the local events in video and can be adapted to size and velocity of the pattern of the event.To evaluate the effectiveness of this method,we use the public Hollywood dataset.In this dataset shot sequences are collected from 32 different Hollywood movies and it includes eight event classes.According to the experiment, the average accuracy rate,the average precision rate,and average recall rate were 0.601,0.530 and 0.566 respectively.The presented results justify the proposed method explicitly improving the average accuracy and average precision compared with other relative approaches.关键词：local space-time interest points;local space-time features;self-organization feature map;event recognition4420|82|0更新时间：2024-05-08
- 摘要：Off-line signature verification is an important form of behavioral biometric identification. We present a method utilizing direction feature and grid feature to tackle the problem. Grid feature has been widely used as one of the mainstream feature extraction approach. The combination of direction feature and grid feature can not only describe the direction and location of the special point, but also record the distribution information of the location of the direction. In order to get features with lower dimensions, principal component analysis is employed to reduce redundant dimensions. In addition, we adopt support vector machines as classifiers for verification process. The proposed strategy is evaluated on the public signatute data bases. Experimental results have demonstrated that the proposed method is effective to improve off-line signature verification accuracy.关键词：off-line signature verification;direction feature;grid feature;combination;support vector machine4429|107|0更新时间：2024-05-08
- 摘要：Stereo image quality assessment is important to the stereo video technologies. Traditional PSNR does not reflect the characteristics of human visual perception and can not be directly apply to the stereo image quality assessment. According to the human visual system of depth perception and focusing on regions of interest (ROI)to stereo image, a new assessing method based on the ROI of texture and depth map is proposed. First, extracting the ROI of texture image and the corresponding depth map. Then, allocating the weighting factors according to the degree of interesting. Finally, assessing the stereo image by applying the weighting factors to the various regions. Experimental results show that this method is consistent with the subjective judgments and can effectively reflect human visual perception characteristics.关键词：visual attention;stereo image quality assessment;depth perception;region of interest (ROI)5115|116|0更新时间：2024-05-08
- 摘要：With the development of the computer technology and the Internet, the application of databases is developing rapidly. However, in the field of computer applications where everything is being continuously optimized, upgraded,and integrated, it is hard to reuse data in a heterogeneous environment.This forms information islands. Based on XML,which is the carrier of heterogeneous data for data exchange, adopting the J2EE and Dom4j technology, we solve the problem of data exchange for relational databases in a heterogeneous environment effectively. By guaranteeing integrity,we provide users with a flexible and low overhead method for heterogeneous data management.关键词：heterogeneous data;data exchange;extensible markup language(XML);relational database3976|61|0更新时间：2024-05-08
- 摘要：The temporal and inter-view correlation of multi-view videos (MVV)changes not only from sequence to sequence but also from frame to frame. Therefore, the fixed prediction structure of the Joint Multi-View Video Coding (JMVC)has difficulties coping with various characteristics of different MVV sequences. In this paper, a flexible prediction structure for multi-view video coding (FPS_MVC)is proposed based on the analysis of the hierarchical B picture prediction structure and the temporal correlation and inter-view correlation between the current frame and its references.Its temporal level (TL)and the temporal and inter-view correlation of the MVV sequence decide whether the inter-view prediction of the current encoded frame is adopted. Experimental results show that the proposed FPS_MVC outperforms Scalable Prediction Structure (SPS_MVC)on reducing computation complexity, improving random access (RA)ability, and decreasing the decode picture buffer (DPB)size by about 45%, 28%, and 46% on average, while maintaining almost the same coding efficiency.关键词：multi-view video coding;prediction structure;temporal and inter-view correlation;random access;decode picture buffer4721|69|0更新时间：2024-05-08
- 摘要：One cross-layer feedback based on video coding scheme is proposed in this paper, which applies both adaptive modulation coding (AMC)and automatic repeat request (ARQ)techniques to provide an adaptive wireless video coding solution. Within this method, timely channel-aware can be obtained by the feedback information from the PHY layer, and then the video encoder at application (APP)layer can take accurate rate control to adapt the channel’s bandwidth varying. Experimental results showed that, compared to the existed scheme without feedback and channel-aware, the proposed method can get a significant improvement in Peak Signal-to-Noise Ratio (PSNR)for wireless video transmission.关键词：wireless video transmission;adaptive coding;cross-layer feedback;adaptive modulation coding (AMC);automatic repeat request (ARQ)4141|63|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