Quadrature mirror filter

About: Quadrature mirror filter is a research topic. Over the lifetime, 955 publications have been published within this topic receiving 28900 citations.

Gait Recognition by Cross Wavelet Transform and Graph Model

Abstract: In this paper, a multi-view gait based human recognition system using the fusion of two kinds of features is proposed. We use cross wavelet transform to extract dynamic feature and bipartite graph model to extract static feature which are coefficients of quadrature mirror filter U+0028 QMF U+0029-graph wavelet filter bank. Feature fusion is done after normalization. For normalization of features, min-max rule is used and mean-variance method is used to find weights for normalized features. Euclidean distance between each feature vector and center of the cluster which is obtained by k-means clustering is used as similarity measure in Bayesian framework. Experiments performed on widely used CASIA B gait database show that, the fusion of these two feature sets preserve discriminant information. We report 99.90 U+0025 average recognition rate.

Multi-scale wavelet modulation

Abstract: The purpose of this paper is to provide the fundamental characterization of a general multidimensional modulation format which utilizes wavelet basis functions as pulse shapes. This type of modulation is particularly effective in distorting channels since it allows for frequency selective processing. In addition to developing the signal format, power spectral density, bandwidth efficiency and error rate performance, an efficient digital implementation utilizing digital filter banks is put forth. Finally, the use of wavelets in communication is seen to generate new pulse shape designs. A practical solution discussed in the paper applies the Smith-Barnwell two-channel filter bank design. >

Perfect reconstruction recursive QMF banks for image subband coding

Abstract: A scheme for implementing a perfect reconstruction QMF (quadrature mirror filter) bank is considered. The scheme is based on a uniform DFT (discrete Fourier transform) L-band analysis filter bank realized using a polyphase decomposition of a lowpass L-th band filter (composed of stable allpass subfilters) followed by an L-band synthesis filter bank (realized using the transpose of the analysis filter bank with the allpass filters replaced by their inverse filters). In the subband coding of finite-length sequences, the instability problem of the inverse filters can be avoided by using the original stable allpass filter in each branch of the synthesis bank with time-reversed input and output. Conditions for perfect reconstruction are derived and illustrated with examples. >

PR-QMF based codes for multipath/multiuser communications

Abstract: One common type of interference in a mobile communications channel is multipath interference. Traditionally, the DSSS/CDMA systems utilize maximal length sequences (m-sequences) as the spreading codes. The m-sequences were chosen because they possess the desired randomness properties and are easily generated via shift registers. The perfect reconstruction quadrature mirror filter (PR-QMF) based spread spectrum codes were proposed as an alternative to the m-sequences. A class of spreading codes based on PR-QMF subband trees is examined. After developing the properties of these codes as they relate to the multipath and multiuser channels, objective and constraint functions are formulated which incorporate these properties. These functions are then used in a multivariable optimization problem to determine the PR-QMF coefficients for use in the subband tree structure. Once the subband tree is complete, the spreading codes can then be extracted for use. It is shown that in the limiting cases, the optimal solutions tend towards TDMA and FDMA like structures.

Complex quadrature mirror filters

Abstract: This paper introduces a technique for the decomposition of a signal into N adjacent sub-bands uniformly spaced in the frequency domain. Each subband is in turn converted into two channels in quadrature subsampled by 1/2 N. It is then shown how the original input signal may be perfectly reconstructed from the 2 N channels, and a practical implementation is suggested. Potential applications to speech compression are discussed.