Return to search

The Study of Aerial Images Enhancement by Bi-dimensional Empirical mode Decomposition

Aerial images had been used in various fields. If the shooting it in conditions of bad weather, that quality was damage. So requires use of image processing methods for improvement. With the growth and progress of technology, image enhancement method is quickly flourishing. That needs of different options for the appropriate treatment.
In this study, image enhancement by the nature of Bi-dimensional empirical mode decomposition (BEMD). Empirical mode decomposition in nonlinear signal processing has excellent results. It can break out some of the base signal and trend. In the BEMD of the signal information
for the edge of the image, the trend term for the light distribution.
Decomposition can be retained the edge features of the image, and
contrast of light and shadow maps make adjustments to improve the
uneven lighting situations. In order to retain the color information, so
the decomposition of the image before the image for color conversion,
will be divided into three¡@degrees of brightness and color images, and
then the brightness of the image for the¡@decomposition of the action.
The use of subjective and objective image evaluation indicators, to
determine the two-dimensional empirical mode decomposition¡@
method with other image enhancement difference. The results show that
the conventional image enhancement has a problem that bright side of the whole image, resulting in reduced edge feature information. However,
BEMD to effectively improve the image problem of uneven brightness
and thoroughly to keep edge features.

Identiferoai:union.ndltd.org:NSYSU/oai:NSYSU:etd-0220112-111808
Date20 February 2012
CreatorsHu , Chun-Ting
Contributorsshiahn-Wern Shyue, none, none, none, Liang-Hwei Lee
PublisherNSYSU
Source SetsNSYSU Electronic Thesis and Dissertation Archive
LanguageCholon
Detected LanguageEnglish
Typetext
Formatapplication/pdf
Sourcehttp://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-0220112-111808
Rightsuser_define, Copyright information available at source archive

Page generated in 0.0013 seconds