Since the publication of the sold-out first edition, NIR spectroscopy has become the standard for rapid, accurate analysis of ingredients and constituents used in the manufacture of food. Near Infrared Technology in the Agricultural and Food Industries, 2nd Edition is an indispensable resource that includes revised and updated chapters and current information from a renowned line-up of international experts in the field.

This important title has been completely revised. New chapters on “implementation,” “industrial applications,” “neural networks,” and a new approach to “qualitative NIR analysis” make this an essential reference for food scientists who wish to stay current.

Since the publication of the first edition, NIR spectroscopy has become a key tool in the precise analysis of food components and the prediction of functionality parameters.

Food technologists at any level will benefit from the breadth of knowledge and helpful spectra provided in this book. Those new to NIR spectroscopy, will find the book to be an excellent primer. Those currently using NIR spectroscopy will find this updated resource essential for gaining a deeper understanding of all aspects of NIR Technology. This is especially true as the future uses of NIR spectroscopy will include grading and classifying materials and organoleptic-type categorization of materials and foods.

Near-Infrared Technology in the Agricultural and Food Industries, Second Edition

The Physics of Near-Infrared Scattering

Donald J. Dahm and Kevin D. Dahm

Introduction
Physical Principles

Absorption, Remission, and Transmission
Reflection from a Surface
Absorption, Remission, and Transmission of a Particle
Formation of a Representative Layer
Reflection in Regions of Higher Absorption

Illustrations of Diffuse Reflection
Theoretical Considerations in Making Measurements

Transmission
Remission

Functional Representation of Absorption and Scatter in a Diffusing Medium

The Kubelka-Munk (K-M) Equation
Using Inherently Nonlinear Functions
Obtaining Absorption and Remission Coefficients from Reflectance Data

Illustrations of K-M Scattering

Scattering from Plastic Particles
K-M Scattering

Summary

Chemical Principles of Near-Infrared Technology

Charles E. Miller

Introduction

Name Dropping
The Size and Speed of NIR

The Spectroscopy of NIR

Light Energy
Vibrational Molecular Energy
Vibrational Spectroscopy—Made Simple
Vibrational Spectroscopy—Made Complicated

Chemical Factors Affecting Vibrational Spectra

The Primary Effect: Functional Group
Secondary Effects

Electronic NIR Spectroscopy
The NIR Complication Factor
NIR Correlation Charts
Conclusion

Data Analysis: Wavelength Selection Methods

William R. Hruschka

Introduction
Calibration, Measurement, and Validation

Calibration
Measurement and Validation
Developing a Calibration Model

Sources of Error

Sampling Error
Reference Method Error
NIR Method Error and Smoothing

Single-Term Linear Regression and the Correlation Plot
Multiterm Linear Regression

Basic Properties
Calculation

The Derivative

Basic Properties
Calculation

The Fourier Transform

Basic Properties
Applications

Other Methods

Component Spectrum Reconstruction
Fast Correlation Transform
Normalizing Spectra
Discriminant Analysis
Neural Networks

Conclusion Appendix

Multivariate Calibration by Data Compression

H. Martens and T. Naes

Introduction

Multivariate Calibration and Validation
Calibration
Validation and Analysis

Linear Prediction and Alternative Ways to Find the Calibration Coefficients

Linear Analytical (Prediction) Equation
Multiple Linear Regression as a Calibration Method to Determine the Calibration Coefficients
Different Classes of Calibration Methods

Statistical Calibration Methods for Multicollinear NIR Data

The General Model Framework
Conventional NIR Calibration Methods: Selecting the “Best” Wavelengths
Hruschka Regression: Selecting the “Best” Calibration Samples
Fourier Transform Regression: Concentrating the NIR Data to the Main Spectral Features
PCR: Concentrating the NIR Data to Their Most Dominant Dimensions
PLSR: Concentrating the NIR Data to Their Most Relevant Dimensions
Calibration Based on Beer’s Model for Mixtures

Analytical Ability and Outlier Detection

Evaluating Analytical Ability
The Importance of Outlier Detection
Analysis of NIR Residuals D. Leverage: Position Relative to the Rest of the Calibration Sample Set
Analysis of the Chemical Residuals
Combined Criteria

Data Pretreatment

Response Linearization
Multiplicative Scatter Correction

Illustration by Artificial Data

Artificial Input Data
Graphical Study of the Input Data
The Effect of Using Insufficient Range of Calibration Samples
Using a Complete Calibration Data Set
PLSR
Outliers
Conclusions

Results for Real Data

The Real Data Sets
Effect of Overfitting
Comparison of Some Calibration Methods
Transformations of NIR Data
Improvements of the PLS Calibration Method

Discussion

The Statistical Calibration Methods
Factors Affecting Choice of Method
Data Pretreatment
Error Detection
Updating

Miscellaneous Topics

Design Is Central in Calibration
Linearity Problems
Other Data Preprocessing Methods
Graphical Interpretation of NIR Calibration Based on Soft Modeling

Conclusions Appendix

Abbreviations and Symbols Appendix
Matrix Operations Illustrated for Multicomponent Analysis

Neural Networks in Near-Infrared Spectroscopy

Claus Borggaard

Introduction
Feed Forward Neural Network Trained by Back-Propagation of Error
An Example of a Feed Forward Network
The Data Flow in the Feed Forward Network
Training the Network — Tuning the Weights
How to Present Data to the Neural Network
Monitoring the Training Process
The Feed Forward Network Used for Classification
Kohonen Self-Organizing Maps
The Architecture of the Kohonen Network
A Training Algorithm for Kohonen Networks
Neural Networks—Advantages and Disadvantages

Disadvantages
Advantages

Conclusions

Near-Infrared Instrumentation

W. F. McClure

Introduction
Components of NIR Systems

Lenses and Mirrors: Collecting Radiation
Radiation Sources
Monochromators
Filters
Detectors

Computerized Spectrophotometry: The COMP/SPEC

General Design
Optomechanical
Optoelectronic
Digital Interface

Performance of the COMP/SPEC

Photometric Noise
Wavelength Precision
Fourier Analysis of Instrument Performance

Software for COMP/SPEC

COMP/SPEC File Structure
Scanning/Analysis
Analytical Software Package
Computerized Spectrophotometric Analytical System

Contemporary Near-Infrared Instrumentation

David L. Wetzel

Introduction
Electronic Wavelength Switching: Diode Array Instruments
Electronic Wavelength Switching: Acousto-Optic Tunable Filter Spectrometer
FT-NIR Instruments
Grating Monochromator Instruments
Interference Filter Instruments
Discrete Source Instruments: LEDs Plus Filters
Special Purpose Instruments
Imaging
Summary

Implementation of Near-Infrared Technology

P. C. Williams

Introduction
Calibration Development

Implementation Steps
Monitoring Instrument Performance

Simplified Approach to the Interpretation of Calibration Efficiency

Accuracy and Precision
Statistical Terms Necessary to the Evaluation of Accuracy and Precision
The Calibration (k) Constants
NIR Reflectance Software
Cross-Validation
Interpretation of PLS Calibrations for Functionality

Variables Affecting Near-Infrared Spectroscopic Analysis

Philip C. Williams and Karl Norris

The Philosophy of Error
Sources of Error in NIR Testing

Factors Associated with the Instrument
Factors Associated with the Sample
Operational Factors
Outliers
Possible Origin of Outliers

Method Development and Implementation of Near-Infrared Spectroscopy in Industrial Manufacturing Support Laboratories

Paul J. Brimmer and Jeffrey W. Hall

Introduction

Laboratory NIR Measurements
Industrial Manufacturing Requirements
Industrial NIR Measurement Requirements

Sampling Requirements

Liquids
Solids
Slurries

Quantitative Analysis

Calibration Development
Spectral Manipulation
Calibration Models
Validation
Calibration Maintenance

Qualitative Analysis

Library Development
Validation
Maintenance

Conclusions

Method Development and Implementation of Near-Infrared Spectroscopy in Industrial Manufacturing Processes

Paul J. Brimmer, Frank A. DeThomas, and Jeffrey W. Hall

Introduction
Process Measurement Requirements

Process Type
Sample Collection and Analysis

Process Sample Interface

Liquids
Solids
Suspensions and Emulsions

Process Instrumentation

Process Analyzer Configurations
NIR Instrumentation
NIR/Process Operator Interface

Quantitative Analysis

Sample Selection
Calibration Modeling Methods
Validation
Maintenance

Qualitative Analysis

Process Requirements

Conclusions

Analytical Application to Fibrous Foods and Commodities

F. E. Barton, II and S. E. Kays

Introduction
Structure and Composition of Forages
The Analysis of Forages
NIR as an Analytical Method
Advantages of the Chemometric Method

Qualitative Near-Infrared Analysis

Howard Mark

Introduction
Data Pretreatments
Mahalanobis Distances
The Polar Qualification System
Principal Components
Soft Independent Modeling of Class Analogies
K-Nearest Neighbors
Correlation Coefficient
Bootstrap Error-Adjusted Single Sample Technique