Scheduled System Maintenance:
On May 6th, single article purchases and IEEE account management will be unavailable from 8:00 AM - 5:00 PM ET (12:00 - 21:00 UTC). We apologize for the inconvenience.
By Topic

Vision, Image and Signal Processing, IEE Proceedings -

Issue 1 • Date Feb 1994

Filter Results

Displaying Results 1 - 12 of 12
  • Optimum classification of non-Gaussian processes using neural networks

    Publication Year: 1994 , Page(s): 56 - 66
    Cited by:  Papers (2)  |  Patents (1)
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (1044 KB)  

    A prerequisite for target detection in synthetic aperture radar and moving target imaging radars is an ability to classify background clutter in an optimal manner. Such radar clutter can frequently be modelled as a correlated nonGaussian process with, for example, Weibull or K statistics. Maximum likelihood (ML) provides an optimum classification scheme but cannot always be formulated when correlations are present. In such circumstances, nonlinear, adaptive filters are required which can learn to classify the clutter types: a role to which neural networks are particularly suited. The authors investigate how closely neural networks can approach optimum classification. To this end, a factorisation technique is presented which aids convergence to the best possible solution obtainable from the training data. The performances of factorised networks are compared with the ML performance and the performances of various intuitive and approximate classification schemes when applied to uncorrelated K distributed images. Furthermore, preliminary results are presented for the classification of correlated processes. It is seen that factorised neural networks can produce an accurate numerical approximation to the ML solution and will thus be of great benefit in radar clutter classification View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Time domain algorithm for the estimation of two sinusoidal frequencies

    Publication Year: 1994 , Page(s): 33 - 38
    Cited by:  Papers (1)
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (340 KB)  

    A computationally simple algorithm is presented that measures one or two frequencies using lookup tables and simple adders. The algorithm is based on a real-time processing of instantaneous frequency and envelope. The algorithm provides a maximum likelihood estimate in the single frequency case. Oversampling is required and the algorithm cannot estimate three or more frequencies. A complete error analysis is presented along with simulation results View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Least-squares spectrum estimation through a neural network-inverse predictor structure

    Publication Year: 1994 , Page(s): 67 - 70
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (212 KB)  

    A method for the spectral estimation of random processes composed of sinusoids in white noise is proposed, based on the least-squares solution of an overdetermined set of linear equations representing the relationship between the power spectrum and the autocorrelation of the process. As operation in real-time is assumed, a problem to be faced is the accurate estimation of the autocorrelation, using few samples of the process. This problem is solved by resorting to an inverse predictor. The proposed approach provides a powerful computational architecture, composed of a neural network and an inverse predictor, that is suited for VLSI fabrication. Numerical examples are presented to illustrate the performance of the proposed method View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Designs of Chebyshev-type complex FIR filters and digital beamformers with linear-phase characteristics

    Publication Year: 1994 , Page(s): 2 - 8
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (472 KB)  

    Based on the theory of optimal polynomial approximation, the authors present a complex alternation theorem which shows the existence of Chebyshev complex FIR filters. According to the theorem, a method for designing Chebyshev-type complex FIR filter and DBFs with linear-phase characteristics is proposed. A zero exchange algorithm and the related procedures are used to iteratively find the best approximation to a variety of desired frequency responses and directivity patterns. Several examples are included to show the efficacy of the designs of FIR filters with multistop-/multipassband responses, DBFs with pencil beams and local low sidelobes, and pattern synthesis with a shaped mainlobe View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Detection of circular arcs for content-based retrieval from an image database

    Publication Year: 1994 , Page(s): 49 - 55
    Cited by:  Patents (4)
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (472 KB)  

    Methods are discussed for producing a summary of geometric content for images from a specific geological application, which can be used for content-based retrieval of interesting images from a large database. Using a data set of 100 images, it is shown that a local are detection scheme provides a more concise and hence a better summary of image content than a conventional edge map, while giving a superior performance View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Superresolution technique with edge-based ringing reduction for passive millimetre-wave images

    Publication Year: 1994 , Page(s): 9 - 12
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (224 KB)  

    Images obtained by passive millimetre-wave systems are heavily blurred and band-limited owing to diffraction. A superresolution algorithm has been developed to improve their spatial resolution beyond the diffraction limit. Initially the spectrum of an image is restored within the passband but the resulting image then contains strong ringing owing to the lack of higher spatial frequencies. These ringing artifacts are then reduced using a piecewise-linear model which identifies and restores sharp edges, so introducing higher spatial frequencies beyond the passband. This method has been applied to synthetised and real millimetre-wave images, but may have applications in other fields such as in magnetic resonance imaging View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • NTSC component separation via Hadamard transform

    Publication Year: 1994 , Page(s): 27 - 32
    Cited by:  Papers (3)
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (420 KB)  

    The authors present a technique for separating the NTSC composite signal into its components (Y, I, Q) by applying an N×N Hadamard transform directly to the composite signal sampled at four times the colour subcarrier. They show that the Y, I, Q component signals are mapped into specific areas in the Hadamard domain. Component separation is achieved by assigning particular transform coefficients to each component signal. Simulation results of component separation applied to five colour images and one black and white test pattern are presented. Both signal to noise ratios and subjective results are presented. The results indicate that no noticeable degradation is expected in typical colour images View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Classified vector quantisation with variable block-size DCT models

    Publication Year: 1994 , Page(s): 39 - 48
    Cited by:  Papers (6)  |  Patents (4)
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (760 KB)  

    The paper describes the classified vector quantisation (CVQ) of an image, based on quadtrees and a classification technique in the discrete cosine transform (DCT) domain. In this scheme, a quadtree is used to segment low-detail regions into variable sized blocks and high-detail regions into uniform 4×4 blocks of various edge and mixed classes. High-detail blocks are classified by an edge-oriented classifier which employs a pattern-matching technique with edge models defined in the normalised DCT domain. The proposed classifier is simple to implement, and efficiently classifies edges to good visual accuracy. The low-detail regions are encoded at very low bit rates with little perceptual degradation, while the encoding of the high-detail regions is performed to achieve a good perceptual quality in the decoded image. Decoded images of high visual quality are obtained for encoding rates between 0.3 and 0.7 bpp View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Modelling of medical magnetic-resonance-imaging signals

    Publication Year: 1994 , Page(s): 71 - 75
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (340 KB)  

    Estimation of images of metabolite concentrations in humans from in vivo magnetic-resonance signals is considered. A 3-dimensional model function is set up in which one dimension pertains to the time domain, and the other two to the reciprocal spatial domain. In the various stages of the estimation fast Fourier transforms and a state-space approach are applied. The computation time can be limited by correcting magnetic-field inhomogeneity and classifying the various parts of the image with the aid of artificial neural networks. In difficult cases spectroscopic prior knowledge is invoked in conjunction with nonlinear least-squares fitting View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Algorithmic engineering in adaptive signal processing: worked examples

    Publication Year: 1994 , Page(s): 19 - 26
    Cited by:  Papers (8)
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (560 KB)  

    Algorithmic engineering provides a rigorous framework for describing and manipulating the type of building blocks commonly used to define parallel algorithms and architectures for digital signal processing. So far, the concept has only been illustrated by means of relatively simple examples relating to the use of QR decomposition (QRD) by Givens rotations for the purposes of adaptive filtering and beamforming. Two more challenging examples are presented that illustrate the use of simple diagrammatic transformations to develop novel algorithms and architectures, and demonstrate the potential power of algorithmic engineering as a formal design technique. The first example constitutes the only known derivation of a modular processing architecture for generalised sidelobe cancellation based on QR decomposition. The second provides a simple derivation of the QRD-based lattice algorithm for multichannel least-squares linear prediction View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Bayesian model selection applied to spatial signal processing

    Publication Year: 1994 , Page(s): 76 - 80
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (308 KB)  

    Inference is applied to the problem of direction of arrival estimation and the determination of the number of sources present in the case of signals impinging on a linear array of sensors. The method is discussed, and the algorithm is found to outperform methods such as AIC and MDL even in the case of very low signal-to-noise ratios and closely spaced sources. The two levels of inference involved in any data analysis problem are introduced and this inference methodology, using the concept of evidence, is then applied to this problem. Simulation results illustrate the power of the method and it is shown that one of the advantages of the present approach is that it is able to handle special cases such as coherent and correlated signals of any functional form, and it does not assume the receivers to be equally spaced View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Fast orthogonal search for array processing and spectrum estimation

    Publication Year: 1994 , Page(s): 13 - 18
    Cited by:  Papers (7)
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (444 KB)  

    An efficient, general purpose modelling algorithm known as fast orthogonal search (FOS) is applied to the problem of estimating the angles of arrival of multiple plane waves incident on an array of sensors and to the related problem of estimating the temporal frequencies of narrowband signals in noise. An iterative version of the algorithm (IFOS), which approaches a minimum in the mean squared error between observed and modelled data, is introduced. A modification to the array steering vector model which is beneficial when the incident signals are coherent is described. FOS and IFOS are compared with MUSIC and root MUSIC in computer simulations View full abstract»

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