 |
Image and Video Compression Fundamentals and International Standards |
|
Introduction to Biological and Artificial Neural Networks for Pattern Recognition |
|
Fuzzy and Neural Pattern Recognition |
|
Digital Image Enhancement
|
|
Image and Video Compression Fundamentals and International Standards |
|
Introduction to Biological and Artificial Neural Networks for Pattern Recognition |
|
Fuzzy and Neural Pattern Recognition |
|
Digital Image Enhancement
|
Majid Rabbani joined the Eastman Kodak Research Laboratories in Rochester, NY, in 1983 and currently serves as head of the Image Coding and Restoration group.
This course will focus on the technical aspects of emerging international industry video standards and will investigate their scope, utility, and performance. The performance of various techniques is further illustrated by use of many monochrome and color image examples in addition to video sequences.
This course will allow you to:
|
Understand the advantages of using digital image and video compression algorithms in various digital imaging related products |
|
Understand fundamental concepts underlying image compression algorithms |
|
Learn noise free symbol coding strategies and characterize their performance bounds |
|
See how color information in a color image is characterized and learn strategies used to compress that information |
|
Learn details of the JPEG algorithm; the international standard for compression of monochrome and color still images. Also, you will be guided in the proper selection of JPEG user-defined parameters |
|
Study technical details of MPEG algorithm, one international standard for compression of motion sequences. |
Part I: Introduction
|
Describe digital image formation |
|
Present examples of existing products to establish the need for digital image and video compression in both storage and transmission applications |
|
Define image compression terminology |
|
Briefly review the scope and the current status of emerging international standards |
Part II: Image Compression Fundamentals
|
Explain constituents of a generalized image compression scheme |
|
Review basic information theory concepts |
|
Describe variable-length and Huffman coding |
|
Describe various quantization strategies |
Part III: The JPEG Baseline Algorithm
|
Present an overview of the scope and the utility of JPEG baseline algorithm |
|
Illustrate the operation of the JPEG baseline through a comprehensive example |
|
Demonstrate the JPEG baseline performance by presenting image examples |
Part IV: Color Image Compression and Extended JPEG
|
Describe the characterization of color in color images |
|
Explain strategies for color image compression and demonstrate via image examples |
|
Explain the JPEG hierarchical modes of operation |
|
Explain the JPEG lossless mode |
Part V: The MPEG Algorithm
|
Present an overview of the scope and the utility of the MPEG algorithm |
|
Motion estimation and compensation |
|
Strategies used for the encoding of the various frame types in MPEG I |
|
Overview of the scope and utility of the MPEG II algorithm |
|
Demonstrate the MPEG I performance by viewing sample motion sequences |
Intended Audience: Primarily scientists, engineers, and technicians with little or no background in digital image and video compression who would like to learn about the emerging international industry standards and their underlying technical concepts. Technical managers who supervise research or development of digital imaging related products will also be interested.
Steven K. Rogers is a professor of electrical and computer engineering at the Air Force Institute of Technology, Wright-Patterson AFB, Dayton, Ohio.
This course provides the background necessary to apply artificial neural networks to problems in machine recognition. The emphasis is on providing a clear and easy-to-understand explanation of the most popular artificial neural network paradigms and how they are applied to engineering problems. Interpretations of current biological, heuristic and mathematical models to allow you to understand current research and potential applications will also be discussed.
|
Gain historical perspectives on research in biological and artificial neural networks |
|
Become aware of how supervised feedforward neural networks can be applied to common engineering problems |
|
Understand how unsupervised feedforward neural networks relate to popular conventional clustering algorithms |
|
Be familiar with applications of popular artificial neural networks |
Part I: Introduction to Biological Neural Networks
|
Define characteristics of neurons and neural networks |
|
Summarize known architectures and characteristics of brains |
|
Describe biological information processing, characteristics and capabilities |
Part II: Overview of Artificial Neural Networks
|
What is new and different as compared with conventional information processing |
|
Historical perspective of biological and artificial network research |
|
Conventional pattern recognition ideas related to neural networks |
Part III: Perception Architectures
|
Single and single layer perceptrons |
|
Multiple layers of perceptrons |
|
Explain backward error propagation learning |
|
Share insights into learning and sample applications |
Part IV: Unsupervised and Hybrid Neural Networks
|
Kohonen unsupervised networks |
|
Hybrid neural networks |
|
Radial basis function networks |
Part V: Feedback Networks and Details on Applications
|
Review popular feedback neural networks |
|
Example of target recognition application |
|
Examples of speech processing and recognition applications |
|
Summarize miscellaneous applications |
|
List various development tools |
|
Present optical implementations |
Intended Audience: Those who need to learn about the use of artificial neural networks to solve pattern recognition problems and those who work with engineers solving machine processing of information. The course will also be of interest to those managing, marketing or supporting projects that might use neural based techniques.
Instructor: James Bezdek is a professor of computer science at the University of Western Florida.
This course covers two technologies: fuzzy models and computational neural networks. It provides information on how both are used in selected pattern recognition applications. Fuzzy sets were introduced by Zadeh in 1965 to represent/manipulate data and information possessing non-statistical uncertainties. Computational neural networks were first discussed by McCullogh and Pitts in 1943 as a means of imitating the power of biological systems for data and information processing.
This video course provides an overview of the major considerations, models and algorithms that can be used for feature analysis, clustering and classifier design. Particular emphasis will be placed on proven methods for applications and system development.
|
Understand theory of fuzzy models and their relationship to probability |
|
Apply design principles for pattern recognition systems |
|
Use algorithms for feature analysis, clustering and classification |
|
Understand clustering and classification with neural network paradigms |
|
Apply the principles of fuzzy neural pattern recognition to boundary analysis, edge detection and segmentation of LANDSAT radar and magnetic resonance images |
Part I: First Ideas About Fuzzy Models
|
Describe uncertainty in models and systems |
|
Explain fuzzy sets and probability |
|
Explain membership functions |
|
Define basic fuzzy set operations |
|
Address common questions about fuzzy sets |
Part II: Numerical Pattern Recognition
|
Define pattern recognition |
|
Describe object and relational data |
|
Explain feature analysis |
|
Show clustering and classification |
|
Discuss system design considerations |
Part III: Clustering and Classification with Fuzzy Models
|
Define partition spaces |
|
Explain hard and fuzzy c-means |
|
Explain image segmentation with c-means |
|
Show fuzzy c-lines: corner detection |
|
Show fuzzy c-shells: boundary analysis |
Part IV: Fuzzy Logic & Clustering Networks
|
Show prototypes in clustering |
|
Describe Kohonen's LVQ and KSO models |
|
Discuss generalized LVQ |
|
Discuss fuzzy LVQ |
|
Explain image segmentation with FLVQ/GLVQ |
Part V: Fuzzy Logic and Classifier Networks
|
Discuss biological neural models |
|
Show computational neural networks |
|
Show feed forward classifier networks |
|
Describe edge detection and feature extraction |
|
Describe image segmentation with NNs |
|
Discuss statistical decision theory |
Intended Audience: this course is designed for people who need to learn about pattern recognition techniques in systems that use sensor data. It will provide an overview of the major considerations, models and algorithms that can be used for feature analysis, clustering and classifier design.
Instructor: Majid Rabbani joined the Eastman Kodak Research Laboratories in Rochester, NY, in 1983 and currently serves as head of the Image Coding and Restoration group.
This course provides an introduction to the basic concepts and techniques used in digital image enhancement, a subfield of digital image processing, that deals with the manipulation of pictures by the computer to increase the pictures usefulness. The focus is on developing basic and state-of-the-art algorithms for the enhancement of monochrome images. The usefulness of each technique is further illustrated by examples of images.
|
Understand the characterization of common degradations in conventional imaging systems and realize the advantages of digital vs. optical image processing |
|
Understand the fundamental concepts underlying the development of image enhancement techniques |
|
Improve a pictures poor contrast by using a variety of adaptive and non-adaptive enhancement techniques |
|
Improve a pictures sharpness by using a variety of space- and frequency-domain adaptive and non-adaptive sharpening techniques |
|
Reduce a pictures graininess and thus improve its appearance or usefulness by using a variety of linear and nonlinear noise reduction algorithms |
|
Observe the effect of different enhancement techniques and their parameter variations on a variety of image examples. |
Part I: Introduction
|
Describe digital image formation and sources of image degradations |
|
Define sampling and quantization |
|
Compare digital image processing to optical image processing |
|
Compare digital image enhancement to digital image restoration Part II: Contrast Manipulation |
|
Define tone scaling |
|
Define contrast stretching |
|
Describe histogram equalization (both continuous- and discrete-case) |
|
Describe histogram modification |
Part III: Contrast Manipulation/Sharpening
|
Explain adaptive histogram equalization |
|
Describe FIR (finite impulse response) linear filtering for image sharpening |
|
Define nonadaptive and adaptive unsharp masking |
|
Describe statistical differencing |
Part IV: Sharpening/Noise Reduction
|
Define high-pass filtering |
|
Explain homomorphic filtering |
|
Describe other transform-based sharpening techniques |
|
Explain linear FIR filtering for noise reduction |
|
Analyze the effect of linear filtering on noise reduction |
Part V: Noise Reduction
|
Compare out-of-range smoothing with directional smoothing |
|
Explain edge and line weights method |
|
Describe noise removal based on use of local statistics |
|
Describe median filtering as a tool for noise reduction |
Intended Audience: Primarily aimed at scientists and engineers with little or no background in digital enhancement who would like to learn about fundamentals and state-of-the-art algorithms in this area. The course will also be of interest to technical managers who supervise research or development in image processing.
Online Comments & Feedback
Online Request for
Information, Quotation and Ordering
 |