By Topic

Vision, Image and Signal Processing, IEE Proceedings -

Issue 5 • Date Oct 2002

Filter Results

Displaying Results 1 - 8 of 8
  • DC coefficient restoration technique and its application to image coding

    Page(s): 272 - 282
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (3242 KB)  

    In the traditional approach of block transform image coding, a large number of bits are allocated to the DC coefficients. A technique called DC coefficient restoration (DCCR) has been proposed to further improve the compression ability of block transform image coding by not transmitting the DC coefficients but estimating them from the transmitted AC coefficients. Images thus generated, however, have inherent errors that degrade the image visual quality. In the paper, a global estimation DCCR scheme is proposed that can eliminate the inherent errors. The scheme estimates all the DC coefficients of the blocks simultaneously by minimising the sum of the energy of all the edge difference vectors of the image. The performance of the global estimation DCCR is evaluated using a mathematical model and experiments. Fast algorithms are also developed for efficient implementation of the proposed scheme. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Nonlinear filtering for phase image denoising

    Page(s): 290 - 296
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (964 KB)  

    The problem of phase image denoising through nonlinear (NL) filtering is addressed. There are various imaging systems in which the phase information is utilised to generate useful imaging data. However, the presence of noise makes difficult to obtain the appropriate phase image. The authors apply NL vector filtering techniques to denoise the complex data from which the phase image is extracted. A study was realised in which several NL filters were applied to a simulated complex image. The effects of filtering were determined through a Monte Carlo simulation in which the image was successively contaminated with six different noise models. The effectiveness of the filters was measured in terms of normalised mean square error, signal-to-noise ratio and the number of eliminated phase residues. Results indicate a significant noise reduction, especially when NL filters based on angular distances are applied to the noisy input. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Comparison of three rough surface classifiers

    Page(s): 263 - 271
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (2560 KB)  

    Texture analysis techniques are used to segment rough surfaces into regions of homogeneous texture. The performance of three rough surface classifiers was assessed and compared. The classifiers differ in their discrimination as well as in their input and computational requirements. Simulation and experiment were used to identify the limitations of the classifiers and to identify which classifier is best suited to a particular task. A series of guidelines for the choice of classifier is presented and justified. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Boundary filter design for multiwavelets

    Page(s): 315 - 320
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (584 KB)  

    When designing boundary filters for multiwavelets, perfect reconstruction and moment conditions are two of the most important criteria to be considered. Since the filter coefficients of multiwavelets are expressed in the form of matrices, the moment conditions for multiwavelets do not relate to the filter coefficients in a simple way as for scalar wavelets. The authors first formulate the moment conditions of multiwavelets and then use them as the criteria for the optimisation of the boundary multifilters. They then apply the proposed optimisation scheme to the design of boundary multifilters for CL and GHM multiwavelets. The optimised boundary filters possess the desired perfect reconstruction property and vanishing moments that allow them to reduce the boundary artefact by up to 75% as compared with the traditional approach. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Fusion algorithm for multisensor images based on discrete multiwavelet transform

    Page(s): 283 - 289
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (1196 KB)  

    The authors review the notion of multiwavelets and describe the use of the discrete multiwavelet transform (DMWT) in image fusion processing. Multiwavelets are extensions from scalar wavelets, and have several advantages in comparison with scalar wavelets. Multiwavelet analysis can offer more precise image analysis than wavelet multiresolution analysis. A novel fusion algorithm is presented for multisensor images based on the discrete multiwavelet transform that can be performed at pixel level. After the registering of source images, a pyramid for each source image can be obtained by applying decomposition with multiwavelets in each level. The multiwavelet decomposition coefficients of the input images are appropriately merged and a new fused image is obtained by reconstructing the fused multiwavelet coefficients. This image fusion algorithm may be used to combine images from multisensors to obtain a single composite with extended information content. The results of experiments indicate that this image fusion algorithm can provide a more satisfactory fusion outcome. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Design of two-channel low-delay IIR nonuniform-division filter banks using L1 error criteria

    Page(s): 304 - 314
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (850 KB)  

    The design of a two-channel nonuniform-division filter (NDF) bank with infinite impulse response (IIR) analysis/synthesis filters and low group delay in the sense of L1 error criteria is considered. The problem formulation results in a nonlinear optimisation problem. Based on a variant of Karmarkar's algorithm, the optimisation problem is solved through a frequency sampling and iterative approximation technique to find the tap coefficients and the reflection coefficients for the numerator and the denominator of the IIR analysis filters. An efficient stabilisation procedure ensures that the reflection coefficients lie in (-1, 1). Simulation results are provided for illustration and comparison. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • General multilayer perceptron demixer scheme for nonlinear blind signal separation

    Page(s): 253 - 262
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (896 KB)  

    A new technique is presented for instantaneous blind signal separation from nonlinear mixtures using a general neural network based demixer scheme. The nonlinear demixer model follows directly from the general mixer model. A general mixer model is described which includes linear mixtures as a special case. In the second part the general framework for a demixer based on a feedforward multilayer perceptron (FMLP) employing a class of continuously differentiable nonlinear functions is presented. A detailed derivation of the learning algorithm used to adapt the demixer's parameters is given. Cost functions based on both maximum entropy (ME) and minimum mutual information (MMI) have been studied. The performance of the new technique was investigated using various experiments derived from the general mixer model and using real-time data. These studies illustrated the superiority and the generality of the new technique compared with existing methods. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Eigenfilter approach for the design of variable fractional delay FIR and all-pass filters

    Page(s): 297 - 303
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (677 KB)  

    An eigenfilter approach is presented for designing 1-D and 2-D variable fractional delay FIR and all-pass filters. First, the coefficients of filters are expressed as a polynomial of the fractional delay parameter. Then, the optimal polynomial coefficients are obtained from the elements of the eigenvector corresponding to the minimum eigenvalue of a real, symmetric and positive definite matrix. Finally, several design examples of 1-D and 2-D variable fractional delay FIR and all-pass filters are used to illustrate the effectiveness of the eigenfilter approach. View full abstract»

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