مرکز تحقیقاتی فناوری اطلاعات و ارتباطات دانشگاه صنعتی شریف

نمایش محتوا

طرحها و پروژه ها/طرح های پژوهش/Matching Pursuit

Project Title:

Matching Pursuit Image and Video Processing

Goals:
Define new and efficient MP feature vectors based on local image texture.
Using MP for better very low bit rate video coding, based on improving some parts of Hybrid-DCT video coding with this regards.

Project Summary:
Matching Pursuit is an algorithm for finding an expansion for energy signals. With use of this algorithm and gabor-wavelet basis, we could expand input images. We work to find appropriate methods to extract feature vectors and very low bit rate-coding of video.

Team Members:
Mehrdad Yaghoobi Vaighan

References:
with M.B.SHAMSOLLAHI, "Using Matching Pursuit Expansion for Image Texture Extraction" 2^nd Biannual Iranian Conference on Machine Vision and Image Processing, pp253-260, Vol. 1, Feb. 13-15, 2003 (In Persian)
with H.R. Rabiee and others, “A New Image Texture Extraction Algorithm Based on Matching Pursuit Gabor Wavelets” IEEE ICASSP2005 (Accepted)
MP Image Feature Extraction Algorithm Source Codes in C++ and Matlab (By Request