By Topic

Image Processing And Its Applications, 1999. Seventh International Conference on (Conf. Publ. No. 465)

Date 1999

Go

Filter Results

Displaying Results 1 - 25 of 83
  • A handwriting understanding environment (HUE) for research in document and handwriting processing

    Page(s): 740 - 744 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (272 KB)  

    Researchers in the fields of document processing and handwriting analysis and recognition increasingly need to construct elaborate software systems to investigate and evaluate new research components such as cursive handwriting algorithms, contextual recognition systems and document parsing tools. Although a number of commercial and public-domain image processing frameworks already exist, for example Khoros, KBVision and the Image Understanding Environment (IUE) none of these systems fully supports the cycle of research, software development, informal and formal evaluation which is typically required by the research community. TABS (The Almost a Blackboard System) was originally developed to address this problem by providing a library of domain-specific image processing and computer vision components within a small, efficient, and easily accessible Tcl Tk script-based software framework. TABS was evaluated over a period of more than a year, and has now been rewritten and extended to produce HUE, a handwriting understanding environment which provides a rapid prototyping environment for document and handwriting research. We review the features of HUE, and illustrate its use with a realistic research evaluation example View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A new modal matching algorithm based on dynamically located sub-regions for point feature correspondence

    Page(s): 716 - 720 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (284 KB)  

    The correspondence problem is often encountered in computer vision. In particular, point feature matching of two images is a critical aspect in: stereo vision, structure-from-motion problems, and moving object tracking in image sequences. Besides, it is one possible approach to define similarity metrics among shapes in content-based retrieval for pictorial databases. This paper proposes an algorithm for reliably matching a large number of point features between two images by means of modal matching, and introduces a new method for the point-matching phase that is the most critical step when dealing with large deformations. The proposed algorithm performs a reliable and exhaustive point-feature association between shapes. The robustness is achieved by limiting risky situations using the principle of locality of associations. The approach does not use any distance based control, therefore it avoids any ad hoc threshold and copes effectively even with very large deformations. Practical experiments have been done for content based image retrieval, and for tracking clouds movements in remote sensing image sequences. In both these applications very promising results have been obtained View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Computerized classification of breast lesions: shape and texture analysis using an artificial neural network

    Page(s): 517 - 521 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (244 KB)  

    In this study we have investigated two groups of features: shape features, extracted from microcalcifications and texture features, extracted from the original regions of interest (ROI), in order to classify early breast cancers which have microcalcifications associated. These features were analyzed using different topologies of the artificial neural network (ANN) multi-layer perceptron (MLP). The performance of the ANN was analyzed with receiver operating characteristic (ROC) methodology View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Performance of different still picture interpolative coding schemes on block matching algorithm

    Page(s): 577 - 580 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (372 KB)  

    Still picture coding techniques such as adaptive interpolative coding are always applied to intraframe image compression in video coding algorithms. Coding fidelity is always utilized in measuring the performance of still picture coding methods. In this paper, a popular motion estimation method known as the block matching algorithm (BMA) is adopted in reconstructing interframe images based on the contents of the intraframe images encoded with different still picture coding techniques. The compatibility of the still picture coding techniques can be reflected from the coding fidelities of interframe images. Experimental results show that adaptive interpolative coding with edge prediction technique has better compatibility with BMA and better visual quality is achieved View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A window-based image processing approach for real-time road traffic analysis

    Page(s): 681 - 685 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (296 KB)  

    Various researchers for a number of years have investigated traffic data collection and analysis. Many techniques have been proposed to speed up the operations and some intelligent approaches have been developed to compensate the effects of lighting, shadows and occlusions. To achieve real-time processing, it is necessary to reduce the amount of data to be processed, so the first question is to determine the suitable location for the key regions or windows. Previously, researchers have either used full frame processing approach, which requires more computing power and thus is not practical for real-time applications, or have used window-based techniques, but defining of windows was done manually, which is not practical in real-world traffic applications. In this paper, an automatic approach is described to measure the size and location of windows required for this purpose. The results demonstrate that this method provides better results than the fixed window size View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A multiresolution based image segmentation

    Page(s): 567 - 571 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (336 KB)  

    Segmentation is an important topic in computer vision and image processing. In this paper we develop a scheme based on multiresolution for segmentation. The multiresolution based segmentation algorithm first segments the image using a known segmentation algorithm at coarse resolution and uses this information to segment images at finer resolutions. In this paper, we sketch a scheme for a multiresolution segmentation algorithm and demonstrate its validity on some real images and compare its performance with the segmented image obtained working at a single resolution View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • High-resolution image reconstruction from multiple low-resolution images

    Page(s): 596 - 600 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (312 KB)  

    In this paper, we demonstrate a digital signal processing (DSP) algorithm for improving spatial resolution of images captured by CMOS cameras. The basic approach is to reconstruct a high resolution (HR) image from a shift-related low resolution (LR) image sequence. The aliasing relationship of Fourier transforms between discrete and continuous images in the frequency domain is used for mapping LR images to a HR image. The method of projection onto convex sets (POCS) is applied to trace the best estimate of pixel matching from the LR images to the reconstructed HR image. Computer simulations and preliminary experimental results have shown that the algorithm works effectively on the application of post-image-captured processing for CMOS cameras. It can also be applied to HR digital image reconstruction, where shift information of the LR image sequence is known View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Full text access may be available. Click article title to sign in or learn about subscription options.
  • A fast adaptive superresolution technique for image restoration

    Page(s): 827 - 831 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (272 KB)  

    The authors propose a fast adaptive method of superresolution using a Wiener filter in conjunction with iteration for projection onto convex sets (POCS). The algorithm is robust and adaptive in noise reduction and has very fast convergence. Moreover the algorithm does not require knowledge of the point spread function (psf) of the imaging system View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Estimating dense motion fields using a local two-component model

    Page(s): 696 - 700 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (632 KB)  

    This paper describes a region based approach to computing dense estimates of 2-D motion fields in image sequences. By dense we mean that we would like to have an estimate of the motion at each pixel location at any point in the sequence. In comparison with alternatives, such as optical flow methods, region based approaches to this problem have a number of advantages. Since they involve correlating regions of pixels, they tend to be more robust in the presence of noise and, if the region window is chosen carefully, they can overcome aperture problems. However, previously these advantages have been limited by the simplicity of the underlying model assumed, i.e. typically a single translational motion. Since in general the latter is only valid within small regions, the gains in robustness can be significantly reduced. We adopt a two-component model of local motion-the change between corresponding regions in adjacent frames is assumed to be caused by the presence of either a single motion or two separate motions. The latter case would typically correspond to a motion boundary caused by a difference in depth, i.e. the region can be considered to consist of two sub-regions corresponding to each motion. In addition, we assume that the global motion field can be described by a set of such regions selected from a quadtree tessellation, i.e. blocks at different spatial resolutions View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Handwritten digits parameterisation for HMM based recognition

    Page(s): 770 - 774 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (224 KB)  

    Handwriting classification or recognition methods based on neural networks (NN) have been extensively studied and they are now well known. This process, which parameterises the geometric structure of the digits as a previous stage to their recognition by the neural network, has the inconvenience of ignoring the sequential character of handwriting. The method proposed explores the improvement introduced in a handwritten recognition system when it incorporates the sequential information of handwriting and the hidden Markov model (HMM) is used as a classifier. The handwritten off-line classifier proposed acquire the handwritten characters by a scanner and after their parameterisation (include noise filtering, binarization, thinning and vectorisation) as a sequence is recognised by the HMM classifier, which provides a good probabilistic representation of sequences having large variations. Different parameterisation techniques are introduced and compared View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Wavelet-based image sequence coding for transmission over HF-channels

    Page(s): 572 - 576 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (324 KB)  

    The authors propose a new combined source-channel coding framework. They introduce a new source codec, which follows a multiresolution concept facilitating a decoding at variable spatial resolutions. This property can be used for a source-encoder controlled UEP. Dependent upon the importance of the actual part of image information, the source encoder tells the channel encoder which grade of error protection is to apply. Furthermore, using this strategy it is possible to adapt the protection strength to the channel state. The new channel codec was designed for digital channels with memory View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Fast image fusion with a Markov random field

    Page(s): 557 - 561 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (268 KB)  

    The author demonstrates that the relatively quick fusion of edge information across a number of registered images (possibly from different sensors) is possible using a Markov random field approach. The speed of the method relies on the fixing the pixel value and only optimising the boundary parameter in the model. This more simple optimisation can be undertaken with the demonstrably less computationally expensive method of iterative conditional modes (ICM). The disadvantage of using the ICM method is that it seeks the nearest local optima. However, for the experiments described here, which used real data, this was not found to be a limitation View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Colour image coding by visually meaningful image patterns

    Page(s): 586 - 590 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (344 KB)  

    Visually meaningful patterns (visual patterns) are a set of predefined image blocks representing information that the human visual system is very sensitive to. A colour image coding technique based on visual patterns is introduced. An image is divided into small blocks each of which is mapped to one of the visual patterns. Each block can then be represented by the pattern and by the colour information of that block. Colour quantisation and entropy coding are applied to achieve efficient coding. High compression ratios (35:1-70:1) have been achieved while maintaining good image quality, comparable to that of state-of-the-art techniques such as JPEG. In addition, the encoded images can be used in their compressed state in content based image retrieval systems based on colour and shape features View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Learning structural change for identification and tracking of vehicles moving in open world scenes

    Page(s): 706 - 710 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (432 KB)  

    This paper describes a system, which is able to track multiple man made objects, typically vehicles moving in a natural open world scene. The system interprets changes in structural features in the scene across consecutive image frames and uses the change in structure to identify and track man made objects (target). No a priori knowledge of any structure within the image is assumed. Differences in statistics between a reference image and the current image generate motion cues that are used to identify regions of interest (ROI). Intensity and edgel information are extracted for the ROI in both the current and reference images. Correlation between the extracted data for each ROI yields an initial recognition of the ROI as either a target or object. This process is repeated on a frame by frame basis generating sets of object and target dynamics. The extracted dynamics are used in conjunction with a high-level reasoning process to solve the frame to frame correspondence process. Identified targets are labelled and tracked, objects that fail both the recognition and correspondence process are removed and no longer processed View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Arithmetic coding with sliding window for control-point based image compression

    Page(s): 601 - 604 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (312 KB)  

    A new sliding-window arithmetic coding method is presented. It is suitable for coding points in image compression applications. In this paper, it is used for gray-level image compression, where both the positions and gray-values of the points have to be coded. The presented method gives a gain of 50-200 percent in compression ratios compared to the Huffman method used by Toivanen (1998). Furthermore, the gain in compression ratios compared to standard arithmetic coding is 3-4 percent View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Wordlength estimation for the enhancement of hand-written word recognition

    Page(s): 750 - 754 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (240 KB)  

    Although the problem of recognising machine printed words has been largely solved using available techniques, no ideal solution has been found for the problem of hand-written word recognition, especially with cursive script. This paper describes a method which can be used to estimate the length of hand-written words. The method shares a number of components with recognition techniques. It does not, however, aim to identify the word or any of its constituent characters; instead, it aims to directly identify the number of letters in the word as supporting information to aid more sophisticated recognition processes. The method of wordlength estimation has potential applications in many areas of text analysis. The work presented here concerns the application of the method in the field of automated bank cheque processing; more specifically, in the recognition of the legal amount field. It is also interesting to note that the method has been tested on two different languages, English and French (i.e. on words extracted from the legal amount field of both British and French cheques), in order to test its generic applicability View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A minimum classification error method for face recognition

    Page(s): 630 - 633 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (216 KB)  

    In this paper, we propose a minimum classification error rate based face recognition system. In our work, the minimum classification error formulation is incorporated into a neural network classifier called a multilayer perceptron. Experimental results show that our system is robust to noisy images and complex background View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Recovering the 3D shape of a road by vision

    Page(s): 686 - 690 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (220 KB)  

    This paper deals with a method designed to recover the 3D geometry of a road by using an on-boarded monocular monochromatic camera. The reconstruction process is able to compute the 3D coordinates of the road axis points. It needs the camera height with respect to the ground as well as the width of the road to be known. It also requires the detection in the image of the road edges. The paper describes with details this process which can be applied whatever the method used to extract the road sides in the image. We have obtained many results on several kinds of roads. The most significant are presented here to illustrate the method View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Handwriting recognition using HMMs and a conservative level building algorithm

    Page(s): 736 - 739 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (240 KB)  

    We describe a method for the recognition of cursively handwritten words using hidden Markov models (HMMs). The modelling methodology used has previously been successfully applied to the recognition of both degraded machine-printed text and hand-printed numerals. This paper addresses the additional problems that result from the greater degree of variability present in cursive script. The impact of this additional variability is reduced by applying various normalisation procedures to the scanned text. A novel conservative level building algorithm is proposed which maintains a list of several lexically plausible match sequences at each stage of the search, rather than decoding using only the most likely state sequence. This permits the use of a lexicon directly within the search procedure rather than as a post-processing step. Results are presented on a single-author database of scanned text. Depending on how many alternatives are maintained in the level building search, top-1 word recognition rates as high as 88% are achieved. These rise to 93% for the correct word being ranked in the top-3 candidates. In addition, the HMM methodology provides a likelihood measure that can be used to identify and thereby reject poorly recognised words. Using this mechanism nearly perfect recognition, (99.2%), can be achieved at the expense of rejecting or deferring a decision on 23% of the data View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Tongue tracking in medical X-ray sequences

    Page(s): 494 - 497 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (348 KB)  

    This payer presents a system for the automated tracking of non-rigid anatomic structures in two-dimensional image sequences. It is applied to X-ray image sequences of the vocal tract. In this application articulatory organs have to be measured to investigate the complex dynamic characteristics of human speech production, with particular interest in the robust boundary detection of non-rigid organs such as the tongue View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Analysis of gamma-ray nuclear resonant absorption (NRA) images for automatic explosives detection

    Page(s): 789 - 793 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (492 KB)  

    A method developed for explosives detection via NRA image analysis is discussed. The results of processing simulated images, as well as experimental NRA-images are presented. The experimental images were obtained by scanning aviation cargo along with mock explosives provided by the US Federal Aviation Administration (FAA) using the NRA system developed at Soreq NRC. NRA is a novel transmission radiography technique that was originally developed at Soreq NRC for airline-baggage and cargo-container inspection View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Page similarity and the Hausdorff distance

    Page(s): 755 - 759 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (276 KB)  

    The VideoWriter is a real-time system that uses digitized images from a computer-controlled video camera to find the location of pages on a desktop. Once an image of a page is extracted, the system must determine if the page has been previously stored. We investigate the use of the Hausdorff distance metric to compare two-level versions of page images. The Hausdorff distance metric does not require exact correspondence between pels in images. We compare the accuracy of the original metric with several modifications View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Object centred optimisation process for robust multiple object tracking

    Page(s): 691 - 695 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (280 KB)  

    In this paper, image processing technologies are utilised in order to aid tracking of multiple objects. The techniques in the paper are demonstrated for an aircraft tracking application, although they can be used in other moving object tracking applications. The aim of this paper is to present an object centred optimisation method, for a sensor, which provides the tracker with optimised object data in order to achieve the best possible tracking performance in complex scenarios. The advanced object centred optimisation method overcomes the difficulties of traditional systems, which tend to be scene based, and provides the tracker with the best possible view of each cued object by separately processing the data for each object View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Using 3D structure and anisotropic diffusion for object segmentation

    Page(s): 532 - 536 vol.2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (476 KB)  

    We describe an advanced technique that can be employed to carry out the segmentation task. The goal is the development of a system capable of solving the segmentation problem in most situations encountered in video sequences taken from real world scenes. For this aim the presented segmentation scheme comprises different processing steps implemented as independent modules. The core of the system is a module for multiscale image simplification by anisotropic diffusion. The remaining modules are concerned with the subsequent image segmentation of the resulting smoothed images. The mathematical model supporting the implemented algorithms is based on the numerical solution of a system of nonlinear partial differential equations introduced by Perona and Malik (1987). The idea at the heart of this approach is to smooth the image in direction parallel to the object boundaries, inhibiting diffusion across the edges. The goal of this processing step is to enhance edges keeping their correct position, reducing noise and smoothing regions with small intensity variations. The techniques for object segmentation presented are based on image simplification by nonlinear diffusion and subsequent extraction of object masks taking into account disparity or motion fields as additional information. Two different strategies are considered in order to extract masks of physical objects in the scene: depth-driven nonlinear diffusion and edge extraction and enhancement in scale-space followed by edge matching in two different views of the same scene View full abstract»

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