By Topic

Machine Vision, 2009. ICMV '09. Second International Conference on

Date 28-30 Dec. 2009

Filter Results

Displaying Results 1 - 25 of 78
  • [Front cover]

    Page(s): C1
    Save to Project icon | Request Permissions | PDF file iconPDF (6559 KB)  
    Freely Available from IEEE
  • [Title page i]

    Page(s): i
    Save to Project icon | Request Permissions | PDF file iconPDF (62 KB)  
    Freely Available from IEEE
  • [Title page iii]

    Page(s): iii
    Save to Project icon | Request Permissions | PDF file iconPDF (110 KB)  
    Freely Available from IEEE
  • [Copyright notice]

    Page(s): iv
    Save to Project icon | Request Permissions | PDF file iconPDF (119 KB)  
    Freely Available from IEEE
  • Table of contents

    Page(s): v - x
    Save to Project icon | Request Permissions | PDF file iconPDF (162 KB)  
    Freely Available from IEEE
  • Preface

    Page(s): xi
    Save to Project icon | Request Permissions | PDF file iconPDF (67 KB)  
    Freely Available from IEEE
  • Organizing Committee

    Page(s): xii
    Save to Project icon | Request Permissions | PDF file iconPDF (82 KB)  
    Freely Available from IEEE
  • Reviewers

    Page(s): xiii
    Save to Project icon | Request Permissions | PDF file iconPDF (88 KB)  
    Freely Available from IEEE
  • Survey of Stable Clustering for Mobile Ad Hoc Networks

    Page(s): 3 - 7
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (339 KB) |  | HTML iconHTML  

    Clustering is an important research topic for mobile ad hoc networks (MANETs) because clustering makes it possible to guarantee basic levels of system performance, such as throughput and delay, in the presence of both mobility and a large number of mobile terminals. A large variety of approaches for ad hoc clustering have been presented, whereby different approaches typically focus on different performance metrics. In mobile ad hoc networks, the movement of the network nodes may quickly change the topology resulting in the increase of the overhead message in topology maintenance; the clustering schemes for mobile ad hoc networks therefore aim at handling topology maintenance, managing node movement or reducing overhead. This paper presents the reasons for clustering algorithms in ad hoc networks, as well as a short survey of the basic ideas and priorities of existing clustering algorithms. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Binarization of Documents with Complex Backgrounds

    Page(s): 8 - 12
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (558 KB) |  | HTML iconHTML  

    In this paper, a novel heuristic based approach adopted from George D.C. Calvacanti algorithm for removing background noise from all types of images with complex background is presented. In this approach Binarization is done by selecting two threshold values, one for foreground and another for background for the separation. Morphological techniques are used for improving the quality of the resultant image. In addition to this PSNR ratio is calculated for all the images and its variation with respect to the intensity of the background noise is observed. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Improved Kernel Common Vector Method for Face Recognition

    Page(s): 13 - 17
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (369 KB) |  | HTML iconHTML  

    The common vector (CV) method is a linear subspace classifier for datasets, such as those arising in image and word recognition. In this approach, a class subspace is modeled from the common features of all samples in the corresponding class. Since the class subspace are modeled as a separate subspace for each class in feature domain, there is overlapping between these subspaces and there is loss of information in the common vector of a class which in turn reduces the recognition performance. In CV method the followed criterion considers only the class scatter matrices. Thus the neglecting of the influence of neighboring classes in CV method also reduces the recognition performance. Generally a linear subspace classifier fails to extract the non-linear features of samples which describe the complexity of face image due to illumination, facial expressions and pose variations. In this paper, we propose a new method called ¿improved kernel common vector method¿ which solves the above problems by means of its appealing properties. First, the introduced between-class and within-class scatter matrices consider the neighboring classes and covariance of a class and this makes the obtained common vector has more discriminant information which increases the recognition performance. Second like all kernel methods, it handles non-linearity in a disciplined manner which extracts the non-linear features of samples representing the complexity of face images. Experimental results on real time face database demonstrate the promising performance of the proposed methodology. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Compressing of Fingerprint Images by Means of Fractals Feature

    Page(s): 18 - 21
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (483 KB) |  | HTML iconHTML  

    Today one of the most applicable methods of identity evaluation is identifying based on fingerprint. Now days, various methods and strategies for identifying fingerprint among modern identifying methods becomes common and usual; but on of the remarkable points is about the volume of saving huge amount of fingerprint images into databases which is so effective on this issue. In this article, a desirable compressing method for fingerprint images is presented that is by means of fractals' specialties. That is, in addition to introducing of 6 basic families for making samples in fingerprint images and by using rotationally geometrical conversion - which is of stable characteristics of fractal images - the encrypting of appropriate parts is done and starts compressing these images. The results from the above method are relatively improved in comparison with other common methods such as JPEG. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Adaptive Binarization of Ancient Documents

    Page(s): 22 - 26
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1005 KB) |  | HTML iconHTML  

    It is common for libraries to provide public access to historical and ancient document image collections. Such document images to require specialized processing in order to remove background noise and become more legible. In this paper the proposed approach is adapted from the kavallieratov's algorithm for cleaning background noise from the ancient documents by iterative global thresholding and local thresholding technique. Finally the image quality is enhanced by using morphological technique and compared with other methods in the process of cleaning. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Fingerprint Verification Using the Texture of Fingerprint Image

    Page(s): 27 - 31
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (336 KB) |  | HTML iconHTML  

    In this paper, a fingerprint verification method is presented that improves matching accuracy by overcoming the shortcomings of previous methods due to missing some minutiae, non-linear distortions, and rotation and distortion variations. It reduces multi-spectral noise by enhancing a fingerprint image to accurately and reliably determine a reference point and then extract a 129 × 129 block, making the reference point its center. From the 4 co-occurrence matrices four statistical descriptors are computed. Experimental results show that the proposed method is more accurate than other methods the average false acceptance rate (FAR) is 0.62%, the average false rejection rate (FRR) is 0.08%, and the equal error rate (EER) is 0.35%. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A Robust Neural System for Objectionable Image Recognition

    Page(s): 32 - 36
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (603 KB) |  | HTML iconHTML  

    A reliable model for human skin is a significant need for a wide range of computer vision applications ranging from face detection, gesture analysis, content-based image retrieval systems, searching and filtering image content on the web, and to various human computer interaction domains. In this paper, a robust neural model for human skin recognition is first presented. Then, a fully automated neural network based system for recognizing naked people in color images is proposed. The proposed system makes use of a fast and precise neural model, called Multi-level Sigmoidal Neural Network (MSNN). Furthermore, the system exploits four different color models in all their possible representations to precisely extract color features from skin regions. Receiver Operating Characteristics (ROC) curve illustrates that the proposed system outperforms other stat-of-the-art schemes of objectionable image recognition in the context of detection rate and false positive rate. Abundance of experimental results are presented including test images and the ROC curve calculated over a test set, which show stimulating performance of the proposed system. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Multibiometrics Belief Fusion

    Page(s): 37 - 40
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (509 KB) |  | HTML iconHTML  

    This paper proposes a multimodal biometric system through Gaussian mixture model (GMM) for face and ear biometrics with belief fusion of the estimated scores characterized by Gabor responses and the proposed fusion is accomplished by Dempster-Shafer (DS) decision theory. Face and ear images are convolved with Gabor wavelet filters to extracts spatially enhanced Gabor facial features and Gabor ear features. Further, GMM is applied to the high-dimensional Gabor face and Gabor ear responses separately for quantitive measurements. Expectation maximization (EM) algorithm is used to estimate density parameters in GMM. This produces two sets of feature vectors which are then fused using Dempster-Shafer theory. Experiments are conducted on multimodal database containing face and ear images of 400 individuals. It is found that use of Gabor wavelet filters along with GMM and DS theory can provide robust and efficient multimodal fusion strategy. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Image Registration Using Ant Colony and Particle Swarm Hybrid Algorithm Based on Wavelet Transform

    Page(s): 41 - 45
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (554 KB) |  | HTML iconHTML  

    Mutual information based image registration has the advantages of high precision and strong robustness. However, the solution of this registration method is easy to fall into the local extremes. To overcome this problem, in this paper we propose a new optimal algorithm for image registration, which combines ant colony algorithm with particle swarm algorithm based on wavelet transform. Experiment results demonstrate our proposed approach effective. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Multimodality Image Registration Utilizing Ant Colony Algorithm

    Page(s): 49 - 53
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (685 KB) |  | HTML iconHTML  

    Image registration is the determination of a geometrical transformation that aligns points in one image of an object with corresponding points in another image. To find the best transformation function we should optimize the similarity measure. The optimization methods are generally divided into two general classes of global and local methods. The problem with local methods is that they trap in local optima. So, in this paper, we use ant colony algorithm as a global optimization method which is based on real ant behavior. The results of our experiments show the effectiveness and better accuracy for this method rather than local methods. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A Hybrid Particle Swarm Steepest Gradient Algorithm for Elastic Brain Image Registration

    Page(s): 54 - 58
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (844 KB) |  | HTML iconHTML  

    Over the course of a neurosurgical procedure, the brain changes its shape in reaction to mechanical and physiological changes associated with the surgery. Hence the use of elastic registration is required. In this paper, we propose a hybrid particle swarm with gradient descent algorithm named as HPSO to solve the problem of Elastic brain Image Registration. The main idea is to find the best transformation function that aligns two images by maximizing a similarity measure through HPSO. There are two major optimization methods, global and local methods. The basic problem with local methods such as steepest gradient is that they usually trap in a local minimum. However, steepest gradient will usually converge even for poor initial approximation. On the other hand, the basic PSO as a global method is sensitive to its initial values. So, we decide to use the steepest gradient as a starting approximation for the PSO method. The results from our experiments show that this hybrid algorithm, besides its simplicity, provides a robust, accurate and effective way for elastic brain image registration. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A Conditional Selection of Orthogonal Legendre/Chebyshev Polynomials as a Novel Fingerprint Orientation Estimation Smoothing Method

    Page(s): 59 - 63
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (801 KB) |  | HTML iconHTML  

    In this paper, a new approach to fingerprint ridge orientation estimation smoothing by a conditional selection of orthogonal polynomials is proposed. This method can smooth the low coherence and consistency areas of fingerprint OF. Also, it is able to estimate the Orientation Field (OF) for fingerprint areas of no ridge information This method does not need any basis information of Singular Points (SPs). The algorithm uses a consecutive application of filtering-based and model-based orientation smoothing methods. A Gaussian filter has been employed for the former. The latter conditionally employs one of the orthogonal polynomials such as Legendre and Chebyshev type I or II, based on the results of the filtering based stage. The experiments have been conducted on the fingerprint images of FVC2000 DB2_A, FVC2004 DB3_A and DB4_A. The results show coarse ridge orientation estimation improvement even in very poor quality images where the orientation information cannot be clearly extracted. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Human Gait Recognition Based on Dynamic and Static Features Using Generalized Regression Neural Network

    Page(s): 64 - 68
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (282 KB) |  | HTML iconHTML  

    Biometric recognition using the behavioral modality of gait is an emerging research area. This paper describes a method for human gait recognition using generalized regression neural networks. The feature space is composed of a combination of dynamic (time-varying) gait signals and static body-shape parameters, extracted from binary silhouettes obtained after background subtraction from human gait sequences. The inputs to the neural network are obtained by performing discrete cosine transform (DCT) on the feature space, followed by selection of transformed coefficients to construct compact vectors. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Hierarchical Video Data Modeling and Indexing for Virtual Scene Construction

    Page(s): 69 - 73
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (877 KB) |  | HTML iconHTML  

    Rapidly growing quantities of digital video have made video data become a more important role in many applications than ever. Virtual scene construction also uses video materials as a rich resource. In order to manage the video materials like salient objects and scenes, an effective video retrieval system is required. In this paper, we present a video database management system using a hierarchical video model consisting of Video-Scene-Shot-SalientObject. We first propose a hierarchical video data model which presents a hierarchical structuring of video material and a hierarchical annotation of video material structure. Then two effective indexing structures are proposed based on the model, including an indexing tree structure which considers the relationship between salient objects and a semantic indexing structure inspired by the inverted file indexing structure. Finally, an implementation of this model is described and the efficiency of this method is evaluated. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Automatic Web News Extraction Using Blocking Tag

    Page(s): 74 - 78
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (485 KB) |  | HTML iconHTML  

    There are various approaches for Web news extraction, including tree-edit distance approach that needs to assume the existence of Web templates, visual wrapper based approach that requires large training sets and statistical approach whose flexibility is low. In this paper, a blocking tag based Web news extraction approach is proposed, which automatically detects the blocking tags that break the Web page down into functional areas and then analyzes the web page according to the blocking tags to find out the news content. We have implemented the proposed news extraction approach in a news search engine which has been applied in business of an intelligence enterprise. Compared with related work, our approach does not require the Web page templates or large training sets, and the complexity is lower. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Estimation of Translation, Rotation and Large Scale Scaling Based on Multiple Scaling Assumptions

    Page(s): 79 - 83
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (962 KB) |  | HTML iconHTML  

    We will be able to use highly parallel processing environments. This paper proposes a method for estimating translations, rotations and scaling reaching 10 times simultaneously based on the multiple scaling assumptions, and represents its performance with motion estimation experiments. A sector region luminosity correlation is used for estimating motion vectors. The sector region luminosity correlation is robust about the rotation and withstands large motion environments. The proposed method makes the assumptions about the scaling and estimates the motion vectors based on the assumptions. Then it randomly creates the pair of the estimated motion vectors. Next, it selects the proper pair using the pre-assumed scaling factor. The selected pairs are included in the set of reliable motion vector pairs. The reliable motion vector pairs decide the translation, rotation and scaling. With large scaling, it is difficult to estimate the motion using the sector region luminosity correlation. But with the assumptions about the scaling, they can work. Experiments show that the proposed method makes much better correlations between images than SIFT does in 10 times scaling changes. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A Combined KPCA and SVM Method for Basic Emotional Expressions Recognition

    Page(s): 84 - 88
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (456 KB) |  | HTML iconHTML  

    Automatic analysis of facial expression has become a popular research area because of it's many applications in the field of computer vision. This paper presents a hybrid method based on Gabor filter, kernel principle component analysis (KPCA) and support vector machine (SVM) for classification of facial expressions into six basic emotions. At first, Gabor filter bank is applied on input images. Then, the feature reduction technique of KPCA is performed on the outputs of the filter. Finally, SVM is used for classification. The proposed method is tested on the Cohen-Kanade's facial expression images dataset. The results of the proposed method are compared to the ones of the combined principle component analysis (PCA) and SVM classifier. Experimental results show the effectiveness of the proposed method. The average recognition rate of 89.9% is achieved in this work which is higher than 87.3% resulted from a common combined PCA and SVM method. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.