Title page for ETD etd-0124103-142051


Type of Document Dissertation
Author Xiao, Ke
Author's Email Address kxiao@lsu.edu
URN etd-0124103-142051
Title Fractal Compression and Analysis on Remotely Sensed Imagery
Degree Doctor of Philosophy (Ph.D.)
Department Geography and Anthropology
Advisory Committee
Advisor Name Title
Nina Lam Committee Chair
DeWitt Braud Committee Member
Kam-biu Liu Committee Member
Michael Leitner Committee Member
Oscar Huh Committee Member
Greg Guzik Dean's Representative
Keywords
  • GIS
  • fractal
  • image compression
  • image processing
  • remote sensing
  • data analysis
Date of Defense 2003-01-17
Availability unrestricted
Abstract
Remote sensing images contain huge amount of geographical information and reflect the complexity of geographical features and spatial structures. As the means of observing and describing geographical phenomena, the rapid development of remote sensing has provided an enormous amount of geographical information. The massive information is very useful in a variety of applications but the sheer bulk of this information has increased beyond what can be analyzed and used efficiently and effectively. This uneven increase in the technologies of gathering and analyzing information has created difficulties in its storage, transfer, and processing. Fractal geometry provides a means of describing and analyzing the complexity of different geographical features in remotely sensed images. It also provides a more powerful tool to compress the remote sensing data than traditional methods. This study suggests, for the first time, the implementation of this usage of fractals to remotely sensed images.

In this study, based on fractal concepts, compression and decompression algorithms were developed and applied to Landsat TM images of eight study areas with different land cover types; the fidelity and efficiency of the algorithms and their relationship with the spatial complexity of the images were evaluated. Three research hypotheses were tested and the fractal compression was compared with two commonly used compression methods, JPEG and WinZip. The effects of spatial complexity and pixel resolution on the compression rate were also examined.

The results from this study show that the fractal compression method has higher compression rate than JPEG and WinZip. As expected, higher compression rates were obtained from images of lower complexity and from images of lower spatial resolution (larger pixel size). This study shows that in addition to the fractalís use in measuring, describing, and simulating the roughness of landscapes in geography, fractal techniques were useful in remotely sensed image compression. Moreover, the compression technique can be seen as a new method of measuring the diverse landscapes and geographical features. As such, this study has introduced a new and advantageous passageway for fractal applications and their important applications in remote sensing.

Files
  Filename       Size       Approximate Download Time (Hours:Minutes:Seconds) 
 
 28.8 Modem   56K Modem   ISDN (64 Kb)   ISDN (128 Kb)   Higher-speed Access 
  Xiao_dis.pdf 8.74 Mb 00:40:27 00:20:48 00:18:12 00:09:06 00:00:46

Browse All Available ETDs by ( Author | Department )

If you have more questions or technical problems, please Contact LSU-ETD Support.