Speakers & Lecture Titles

 

B.S. Daya Sagar
Processing of remotely sensed data in both spatial and frequency domains has received wide attention. The application of remote sensing in various fields is greatly realized in the last three decades. One of the data derivable from remotely sensed data is a Digital Elevation Model (DEM) that provides rich clues about physiographic constitution of Earth planet, and Earth-like planetary surfaces. Remotely sensed data are available for various phenomena related to terrestrial, lunar, planetary surfaces, and atmospheric phenomena such as clouds in spatiotemporal mode. To address the intertwined topics—like pattern retrieval, pattern analysis,  spatial reasoning, and simulation and modeling for understanding spatiotemporal behaviors of several of terrestrial phenomena and processes that could be acquired through remote sensing mechanisms—various original algorithms and modeling techniques that are mainly based on mathematical morphology (Matheron 1975, Serra 1982), fractal geometry (Mandelbrot 1982), and chaos theory (May 1976) have been developed and their utility has been demonstrated. During this workshop, a series of lectures on theory and applications of mathematical morphology and scaling concepts in addressing those mentioned intertwined topics. The key links that were shown between those topics–would be highlighted in a set of SEVEN lectures.

Sl.No.
TOPIC
01
 Introduction to Mathematical Morphology
02
 Mathematical Morphology in Terrestrial Pattern Retrieval
03
 Mathematical Morphology in Terrestrial Pattern Analysis
  • Terrestrial Surface Characterization: a Quantitative Perspective
  • Size distributions, Spatial Heterogeneity and Scaling Laws
  • Morphological Shape Decomposition: Scale Invariant but Shape Dependent Measures
  • Granulometries, Convexity Measures and Geodesic Spectrum for DEM Analyses
04
 Mathematical Morphology in Geomorphologic Modelling and Simulation
  • Fractal-Skeletal-Based Channel Network Model
  • Synthetic Models to Understand Spatio-Temporal Dynamics of Certain Geo(morpho)logical Processes
05
 Mathematical Morphology in Quantitative Spatial Reasoning and Visualization
06
 Mathematical Morphology in Spatial Interpolations
  • Conversion of Point-Data into Polygonal Map via WSKIZ
  • Visualization of spatiotemporal behavior of discrete maps via generation of recursive median elements
07
 Quantitative Characterization of Complex Porous Phase via Mathematical Morphology and Fractal Geometry

 

Saroj Kumar Meher

Lectures schedule

Sl.No.
TOPIC
01
 Pattern Recognition:
  • Introduction,
  • Overview of different approaches
  • Decision boundaries
  • Discriminant functions (linear and non-linear)
  • Training and test sets
  • Parametric and nonparametric learning
  • Minimum distance classifiers, k-NN rule,
  • Unsupervised learning, basic hierarchical and non-hierarchical clustering algorithms,
  • Dimensionality reduction, similarity measures, feature selection criteria and algorithms, principal components analysis, some applications.
  • Semisupervised and co-training learning algorithms
02
 Intelligent method of PR: Fuzzy Logic and Applications:
  • Brief overview of crisp sets
  • The notion of fuzziness; what, why and when to apply fuzzy set (Probabilistic and possibilistic)
  • Operations on fuzzy sets; fuzzy numbers. Crisp relations, fuzzy relations, Max_-composition of fuzzy relation; Max_-transitive closure; probability measures of fuzzy events; fuzzy expected value. Approximate reasoning, different methods of rule aggregation and defuzzification
03
 Intelligent method of PR: Neural Networks and applications:
  • Introduction to neural networks, threshold logic, circuit realization. Introduction to biological neural networks, significance of massive parallelism. Perceptron, perceptron learning rule and its convergence, multilayered perceptron, learning algorithms, function approximation, generalization, VC-dimension.
  • Neuro-fuzzy computing and other hybridization, independent component analysis.
04
 Granular computing based pattern recognition:
  • Fuzzy granulation
  • Rough granulation
  • Neural granulation and
  • Hybridization of the above