|Wavelets and subband Coding|
|Image and multidimensional signal processing|
|Multimedia coding and communication|
|Multimedia delivery over Wireless Internet|
I'm a student member of
|Institute of Electrical and Electronics
|Society for Industrial and Applied Mathematics
I'm a reviewer for:
|IEEE Trans. on Multimedia|
|IEEE Trans. on Signal Processing|
|IEEE Trans. on Image Processing|
|IEEE Trans. on Circuits and Systems|
|Journal of Visual Communication and Image
|Conferences, such as GlobeCom, ICASSP, ICME, and
|Nonsubsampled Contourlet Transform and
The Contourlet transform is an efficient directional multiresolution
Owing to downsampling, the Contourlet transform is not shift-invariant. But shift-invariance is required in some
image processing applications, such
as the edge detections, contour detection and image segmentation. We work on
the nonsubsampled Contourlet and its applications on image analysis. This
work is in collaboration with
A. da Cunha
and Jason Laska.
Here are some examples for image enhancement
by nonsubsampled Contourlet transform.
|Multichannel Multivariate Deconvolution and
Over the last
decade, the theory and applications of multichannel deconvolution have
grown rapidly, such as multichannel image deconvolution, channel
equalization for multiple antennas and polarimetric calibration of radars.
We present a general framework for multichannel multivariate exact
deconvolution with FIR blurring filters and FIR reconstruction filters
using the algebraic geometry.
|Multidimensional Filter Banks and Wavelets|
We work on the characterization and design of orthogonal filter banks in
any dimensions. Traditional design methods for one-dimensional orthogonal
filter banks cannot be extended to higher dimensions directly, since there
is no multidimensional (MD) spectral factorization theorem. Due to
difficulty and complexity, there are few design results for MD orthogonal
filter banks. We work on the characterization and efficient design methods
of MD orthogonal filter banks.
|Audio Coding and Streaming over the Internet and Wireless|
Streaming high-fidelity audio over the Internet or the wireless networks
is a challenging task because the channel presents packet loss or bit
errors. These errors may have severe adverse effect on decompressing the
received audio bitstream or even completely crash the decoder. To solve this
problem, we work on error resilience in the audio coder in conjunction with
error protection for scalable audio streaming.
We explore new multiresolution directional image analysis tools and apply
them to image processing applications such as edge characterization, contour
detection, image denoising, image restoration and image enhancement.
Home | About Me | Courses | Research | Publication | People
This site was last updated