Radu Matei

Bio: Radu Matei is an academic researcher. The author has contributed to research in topics: Prototype filter & Network synthesis filters. The author has an hindex of 2, co-authored 7 publications receiving 12 citations.

Multi-directional filters designed from 1D prototypes

Abstract: In this paper a design method based on spectral transformations is proposed for a particular class of 2D IIR filters, namely multi-directional filters. The design starts from an analog prototype with specified parameters. Applying an appropriate frequency transformation to the 1D transfer function, the desired 2D filter is directly obtained in a factorized form. The approach is mainly analytical but also uses numerical approximations and is simple, efficient and versatile. For two-directional filters, an example is given of extracting lines with two different orientations from a test image.

Efficient design of adjustable circular IIR 2D filters

Abstract: This paper proposes an analytical design method for 2D zero-phase recursive filters with a circularly-symmetric frequency response. The design is based on a maximally-flat low-pass prototype filter, whose transfer function is transformed using a specific 1D to 2D frequency mapping. An efficient pre-warping is applied before using the bilinear transform, which allows to obtain filters having a central frequency close to the frequency plane margins, and still with an accurate circular shape, without any distortions. Finally the general filter matrices are obtained. The filter is adjustable in the sense that its bandwidth is specified by a parameter which appears explicitly in the filter coefficients.

Analytical Design of Elliptically-Shaped 2D Recursive Filters

Abstract: This paper presents an analytic design technique for 2D IIR filters with elliptical symmetry, which have useful applications in image processing. The design is based on efficient elliptic digital filters, regarded as 1D prototypes, to which specific complex frequency transformations are applied; this allows to obtain directly a factored form of the transfer function for the 2D elliptically-shaped filter. The design procedure uses accurate approximations, but no global optimization algorithm. Finally the 2D filter matrices are obtained. The filter is adjustable, its coefficients depending explicitly on the specified orientation and bandwidth. Another advantage is versatility, since the design need not be resumed each time from the start for various specifications. The designed 2D filters have an accurate elliptical shape with low distortions even close to the frequency plane margins, being efficient, of high selectivity and low order.

Design Procedure for 2D Gaussian Anisotropic Filters

Abstract: This paper proposes an analytical design method for a class of 2D anisotropic filters, with a directional frequency response, specified by a given orientation angle. These filters have a Gaussian shape in a vertical section and a relatively straight, linear shape in a horizontal section, parallel with the frequency plane. The design starts from a Gaussian prototype with a specified selectivity, to which successive frequency transformations are applied, yielding the transfer function in matrix form of the 2D anisotropic filter with a desired selectivity and orientation. The filter is quite efficient and adjustable, being specified by parameters which appear explicitly in the filter matrices, and allow for adjusting independently their selectivity and orientation. The resulted filters have an accurate shape of the frequency response, high directional selectivity and relatively low order. Moreover, an advantage of this analytical approach is that the filter transfer function results in a parametric, factored form, which allows efficient implementation and easy tunability.

Design technique for two-directional recursive filters

Abstract: This work approaches the design of a class of 2D recursive filters, with a two-directional orientation-selective frequency response. We propose two design methods based on a given 1D digital prototype filter and frequency transformations, one completely analytical, including the bilinear transform, and another which uses a numerical approximation stage. These filters may find useful applications in image processing, like detecting lines with a given orientation from an image, as we show through simulation results. The resulted filters are very efficient, being at the same time of low complexity and relatively high selectivity. Simulation results are provided which show their image filtering capabilities.

Circular IIR Filter Design and Applications in Biomedical Image Analysis

Abstract: This work presents an analytical design method for a category of two-dimensional IIR filters, specifically with circular symmetry, which have useful applications in image processing. The design is based on efficient digital elliptic filters, used as 1D prototypes, to which some complex frequency transformations are applied; this allows to obtain directly a factored form of the transfer function for the 2D circularly-symmetric filter. The design procedure involves some accurate approximations, but no global optimization algorithm. As a result, the 2D filter matrices are derived. The filter has the advantage of being adjustable, as its bandwidth is determined by a parameter which appears in the filter coefficients. The designed 2D filters have an accurate circular shape, exhibiting very low distortions even near the margins of the frequency plane, and are efficient, of high selectivity and relatively low order. The designed filters can be used as a bank of circular filters, useful in a multi-resolution analysis of an image. To prove their filtering capabilities, simulation results are provided for circular filtering of a biomedical image, namely a retina angiogram.

Analytic Design of Uniform Circular Filter Banks

Abstract: This work proposes an analytic design method for a particular class of 2D filter banks, with circular shape in the frequency plane and a maximally-flat characteristic. The circular filters comprising the filter bank result starting from a zerophase prototype by applying specific frequency mappings; they are efficient, of relatively low order and good selectivity. Their transfer functions can be expressed in a closed form and depend explicitly on a parameter given by specified number of filters, e.g. in the case of an uniform circular filter bank. A design example is provided for imposed specification and the designed filter bank is tested in filtering a given texture image.

Multi-directional filters designed from 1D prototypes

Abstract: In this paper a design method based on spectral transformations is proposed for a particular class of 2D IIR filters, namely multi-directional filters. The design starts from an analog prototype with specified parameters. Applying an appropriate frequency transformation to the 1D transfer function, the desired 2D filter is directly obtained in a factorized form. The approach is mainly analytical but also uses numerical approximations and is simple, efficient and versatile. For two-directional filters, an example is given of extracting lines with two different orientations from a test image.

2D Lattice FIR digital filter with alternate delays: Minimal circuit and state-space realization

Abstract: This paper proposes a new two dimensional (2D) lattice structured digital filter, having a minimal number of delay elements. This filter, besides having a minimal number of delay elements, also has an absolutely minimal state-space vector. The new finite impulse response (FIR) digital filter is characterized by lattice structure having alternate delay element orientation. It is worth mentioning that the transfer function coefficients of the proposed 2D filter, are complements of the classic lattice filter. The results of this paper are directly applicable to one-dimensional (1D) lattice digital filters. First and second degree 2D examples are provided to show the features of the circuit and state-space realization structures.