Courses

• 1051.553 Remote Sensing Systems (To be offered next in Winter Quarter 2012-13)
  
  Prerequisites: 1051-370, 1051-452, 1052-453, or permission of instructor
  Credit Hours: 4
  Required Text: Remote Sensing: The Image Chain Approach, 2nd Edition, John Schott, Oxford Publishing, 2007
  Course Description:This course develops knowledge and understanding of the design and analysis of optical remote sensing systems for Earth remote sensing. Building on general imaging fundamentals learned earlier in their program, students will learn domain specific tools and techniques for analyzing airborne and satellite sensor systems for the optical spectral imaging of Earth. Through a combination of classroom and laboratory experiences, students will learn about the propagation of photons and signals from the Sun through the formation of a digital image. The course will emphasize a linear systems modeling perspective and provide the students the background to understand, model, and predict remote sensing imaging system performance.
  
  Right-Click here to download the VNIR image
  Right-Click here to download the SWIR image
  
  
• 1051.753 Special Topics: Synthetic Aperture Radar Imaging (Next offered in Spring 2012)
  
  Prerequisites: Graduate standing, Linear Image Math and Programming
  Credit Hours: 4
  Text: Spotlight-Mode Synthetic Aperture Radar: A Signal Processing Approach by Jakowatz, 1996
  Course Objectives: This course covers the history, fundamental principles, and system requirements for high-resolution spotlight synthetic aperture radar (SAR) imaging. Topics included are synthetic aperture radar concepts, surface and volume scattering, linear models of SAR signal history, image formation and processing algorithms, image quality requirements and SAR system performance, phase errors, and autofocus algorithms. The emphasis in the course is placed on understanding the fundamental principles of spotlight SAR imaging and upon the factors that drive the image quality of SAR products. Along the way, a variety of remote sensing and linear systems theory will be employed to provide specific insight into the following system performance metrics: image size, area rate, resolution, impulse response, noise equivalent backscatter, multiplicative and additive noise, residual quadratic phase error, dynamic range, depth of focus, geometric distortion, oversampling, and signal-to-noise ratio
  
  Homework 2 Test Data
  


• 1051.763 Remote Sensing: Spectral Image Analysis (Spring Quarter 2012-13)
  
  Prerequisites: 1051-719, 1051-762, or permission of instructor
  Credit Hours: 4
  Required Text: Remote Sensing: The Image Chain Approach, 2nd Edition, John Schott, Oxford Publishing, 2007
  Course Description:This course is focused on analysis of high dimensional remotely sensed data sets. A review of the properties of matter that control the spectral nature of reflected and emitted energy is followed by analysis of image noise characterization and mitigation, radiometric calibration, atmospheric compensation and dimensionality characterization and reduction. The remainder of the course focuses on spectral image analysis algorithms employing the three conceptual approaches to characterizing the data for image segmentation, subpixel detection and pixel unmixing, and target detection. The analysis methods include treatment of signal processing theory and application and incorporation of physics based algorithms into spectral image analysis. Also offered online.
  
  
• 1051.784 Pattern Recognition (next taught by Dr. Kerekes in Spring 2013)
  
  Prerequisites: Graduate standing, Linear Image Math and Programming
  Credit Hours: 4
  Text: Statistical Pattern Recognition, 2nd Edition, 2002, by Webb
  Text: Pattern Recognition and Machine Learning, 2006, by Bishop
  Text: Pattern Recognition, 4th edition, 2009, by Theodoridis and Koutroumbas
  Course Objectives: This course develops a fundamental understanding of adaptive pattern recognition and a basic working knowledge of techniques that can be used in a broad range of applications. Inherent in adaptive patter recognition is the ability of the system to learn by supervised or unsupervised training, or by competition within a changing course environment. The effectiveness of a system depends upon its structure, adaptive properties and specifics of the application. Particular structures that are developed and analyzed include statistical pattern recognition, clustering, multilayered perceptrons (with a variety of weight training algorithms), and evolutionary learning systems. The goal is to gain both a fundamental and working knowledge of each kind of system and the ability to make a good selection when faced with real applications.
  Click here for the course syllabus from 2009
  
  MISI Image Data File
  MISI Image ENVI Header
  MISI Image JPEG
  
  Final Project Training Data
  Final Project Training Classes
  Final Project Testing Data
  
  

Return to Dr. Kerekes' Home Page

Last updated: 12 March 2012