|
|
|
Full CV
- I have a wide range of interests, including land-use and
land-cover modeling/mapping, forest type mapping, urban growth prediction and
mapping, assessment and monitoring of land degradation and
desertification, coastal environment management information system,
soil salinization and nutrient depletion modeling, and coastal
suspended sediments and suspended solids mapping. My work has
involved analysis of remotely sensed data, geographic information
system, geostatistical modeling, data mining, pattern recognition,
and geospatial analysis techniques.
- Additionally, my expertise in geospatial techniques, statistical
modeling, and signal processing has led to development of spatial
and frequency based algorithms that identify complex spatial
features, objects, and classes. Some of the geospatial algorithms
that I have been developing and exploring includes spatial
coccurrence matrix, spatial autocorrelation (Moran,s I and Geary's
C), fractal analysis method (e.g., Isarithm approach, Triangular
prism approach, Variogram approach), Lacunarity analysis techniques
(e.g, binary approach, differential box counting method, other gray
scale methods), G index, and Fourier transform approach. I have been
examining the effectiveness of the above geospatial approaches in
identifying urban land use and land cover classes and mapping
coastal features.
- I also explore spatial distribution, dispersion, orientation,
pattern, and association of socio-economic functional units using
central tendency and dispersion approaches, quadrat method, nearest
neighbor method, geographically weighted regression approach, and
spatial analysis on a network (SANET, 2001) using network K-function
method and network cross K-function method.
- Most recently, my research efforts have focused on geospatial
and frequency based multi-scale multi-decomposition techniques for
spatial data mining and pattern recognition. I have developed a new
wavelet-based classification framework and a number of operational
algorithms using Haar wavelets, Daubechies wavelets, and Coieflets
approaches to identify complex land-use and land-cover classes
accurately.
|
