Lab View  
HOME PEOPLE RESEARCH PUBLICATIONS COURSES SPONSORS    

RESEARCH  HIGHLIGHTS

     Projects

     Demos

     Software

     Datasets

  ABOUT US

     Seminars

     Lab

     News

 

  LINKS

Igert

Vision

Ece

Ucsb

Melvyl

 

 

 

 

Project: Quadrature Mirror Filter Bank Design

PEOPLE

 Rajeev Gandhi, S.K.Mitra

OBJECTIVE

One-dimensional quadrature-mirror filter (QMF) banks have been used in the sub-band coding of speech and images. Ideally, these filter banks should be designed to have the perfect reconstruction property in which case the output signal is a delayed replica of the input signal in the absence of quantization of the sub-band signals and the filters. In many applications, it is also desired that the individual filters in the filter bank have linear phase.

In one approach to the design of a two-band 1-D QMF bank, the analysis and synthesis filters are chosen to eliminate aliasing and have linear phase. The design then involves selecting the coefficients of one of the analysis filters by iteratively minimizing the residual amplitude distortion. However due to the large  this project, two new methods for the computationally efficient design of two-band QMF banks based on the frequency-sampling approach have been developed. In the new approach, the number of parameters to be optimized is reduced significantly, thereby leading to a faster design of the QMF banks. The characteristics of the filters obtained are comparable to those of some of the existing QMF banks, in terms of the overall amplitude distortion and minimum stop-band attenuation. A paper describing the results of this investigation was presented at an international conference. In the second part of this project, another new technique for the design of two-band perfect reconstruction filter banks has been developed. This new method is generic, and can be applied to the design of both orthogonal as well as linear-phase bi-orthogonal perfect reconstruction filter banks. Unlike previous design techniques where the perfect reconstruction property was structurally imposed, the new method starts with trivial filters that yield perfect reconstruction and the lengths of the filters are subsequently increased to improve their magnitude responses without sacrificing the linear-phase/orthogonality or the perfect reconstruction property. A paper containing the design method and simulation results was presented at an international conference.

 

PUBLICATIONS

These materials are presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each authors copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.

  1. R. Gandhi, S.K. Mitra, "An efficient top-down approach for the design of tree-structured orthonormal filter banks," Proceedings of 1999 Asilomar Conference, Pacific Grove, CA, USA, Oct. 1999, p.p.1627-1631.

  2. R. Gandhi, S.K. Mitra, "Design of optimal orthogonal tree-structured filter banks, " 42nd Midwest Symposium on Circuits and Systems, Las Cruces, NM, USA, 8-11 August 1999,  p.p.1057-60.

  3. R. Gandhi and S. K. Mitra, "Design of two-channel low delay perfect reconstruction filter banks,"  Proceedings of 1998 Asilomar Conference on Signals, Systems and Computers , Pacific Grove, CA, USA, November 1998, p.p.1655-1699.

  4. R. Gandhi and S. K. Mitra, "A new approach to the design of two-channel perfect reconstruction filter banks," Proc. 5th IEEE International Conference on Electronics, Circuits, and Systems, Lisbon, Portugal, September 1998, pp. 405-408.

  5. R. Gandhi and S. K. Mitra, "A computationally efficient design of two-band QMF banks based on the frequency-sampling approach," Proceedings of the 1998 IEEE International Symposium on Circuits and Systems, ISCAS '98 Monterey, CA, USA, 31 May-3 June 1998.  pp.421-424 

 

 

If you have any comments or questions, please send them to web administrator.

Image Processing and Vision Research Lab, Electrical and Computer Engineering 

Department,  University of California at Santa Barbara, Santa Barbara, CA 93106.