ABSTRACT:
Airborne LIDAR has become commercially used for many environmental, engineering and civil
applications, and can provide accurate data for topographic surfaces and non-terrain objects. Feature
extraction is one of the important applications in the field of photogrammetry. This application is
used for many civil and military applications - such as photo-interpretation, vegetation, forest
monitoring, traffic and transportation development and urban planning. LIDAR Data can have
much more dense point spacing than is typically derived from photogrammetry and therefore proper
handling of the data is needed to optimal feature extraction.
Wavelet transform techniques are widely used as powerful tools in many image processing
applications such as de-noising and compression. The discrete wavelet transform is particularly
suitable in de-noising and filtering problems. The properties of the wavelet transform, such as
having compact support, space and frequency localization, a wide variety of base functions, denoising,
thresholding, and multi-resolution analysis, are the main motivations for testing the wavelet
transform as an estimation technique for feature extraction from LIDAR data.
The major objective of this research is to test the efficiency of wavelet transform in analysing
LIDAR data for feature extraction applications. Wavelet transform succeeded in detecting positions
of sudden changes of the geometrical or physical content of the images from LIDAR data.