Spectroscopy, Spectral Discrimination and Genetic Algorithms
So, what is object parameter estimation using spectral data, i.e. spectroscopy? What is spectral discrimination? Do they have something in common? Can a machine learning approach help to tackle both problem?
This textbook course answers all these questions and more! The textbook book course presents not only the basic theoretical principles of spectroscopy, spectral matching, labeling and discrimination, but also a new novel method, the kstep methodology, that automates the entire process. Both for object parameter estimation and spectral discrimination!
A machine learning approach is incorporated to achieve the full automation; the simple genetic algorithm.
For all these topics, extensive measurements were collected and experiments were performed in order to prove the concept.
Spectral measurements of different varieties of plants (vetch and lentil) were used to showcase the subtle spectral discrimination concept.
Regarding the parameter estimation, soil spectral measurements were taken along with chemical analysis to quantify the soil organic matter.
A textbook to get started with automatic parameter estimation and discrimination using hyperspectral data
Course curriculum

01
Introduction
Show details Lesson 1: Introducing the contents of the textbook FREE TRIAL
 Lesson 2: Object Parameter Estimation workflow or Spectroscopy FREE TRIAL
 Lesson 3: Step by Step Analysis
 Lesson 4: Spectral Discrimination

02
Background: Spectral PreProcessing Algorithms
Show details Lesson 5: Spectral PreProcessing Algorithms (SPPAs) FREE TRIAL
 Lesson 6: Smoothing
 Lesson 7: Vector Normalization
 Lesson 8: Value Normalization
 Lesson 9: Discrete Fourier Transform
 Lesson 10: Logarithm Transform
 Lesson 11: KubelkaMunck Transformation
 Lesson 12: N Order Square Root Transformation
 Lesson 13: Derivatives
 Lesson 14: Continuum Removal
 Lesson 15: Band Depth

03
Background: Spectral Matching & Labeling and Similarity Measures
Show details Lesson 16: Spectral Matching and Labeling FREE TRIAL
 Lesson 17: Similarity Measures
 Lesson 18: Spectral Angle Mapper
 Lesson 19: Cross Correlation
 Lesson 20: Spectral Information Divergence
 Lesson 21: SID – SAM Mixed Measure
 Lesson 22: Continuum Intact Continuum Removed (CICR)

04
Background: Regression Algorithms
Show details Lesson 23: Main Regression Algorithms Used
 Lesson 24: Two Band Normalized Difference Regression (NDR)
 Lesson 25: Multiple Linear Regression (MLR)
 Lesson 26: Partial Least Squares Regression (PLSR)

05
Background: Machine Learning with the Simple Genetic Algorithm
Show details Lesson 27: The Genetic Algorithm Concept

06
Methodological Approach
Show details Lesson 28: Object Parameter Estimation (Part 1)
 Lesson 29: Object Parameter Estimation (Part 2)
 Lesson 30: Object Parameter Estimation (Part 3)
 Lesson 31: Object Spectral Discrimination

07
Use case: Spectral Discrimination of Plant Species
Show details Lesson 32: Data and experiment
 Lesson 33: Experiment Results

08
Use case: Soil Parameter Estimation
Show details Lesson 34: Data and experiment FREE TRIAL
 Lesson 35: Experiment Results for NDR
 Lesson 36: Experiment Results for MLR
 Lesson 37: Experiment Results for PLSR

09
Conclusions
Show details Lesson 38: Conclusions on Spectral Discrimination Use Case
 Lesson 39: Conclusions on Parameter Estimation Use Case
 Lesson 40: Overall Conclusions (Part 1)
 Lesson 41: Overall Conclusions (Part 2)

10
Bibliography
Show details References
What will you learn?

Spectral PreProcessing Algorithms

Simple Genetic Algorithm

Spectral Matching, Labeling, Discrimination

Regression Algorithms

Spectroscopy

Spectral Similarity Measures
Any prerequisites?

Practically none. Just some basic understanding of remote sensing
Student Profile?

Under/post graduate students

Professionals and Companies

Master students and PhD candidates

Researchers and Academics