Summary

This 3-day course provides an introduction to digital signal processing (DSP) tools and techniques used to analyze signals and design systems in one- and two-dimensions. DSP operations such as spectral estimation, frequency selective filtering, and image reconstruction from projections are treated. In contrast to typical DSP courses that focus on mathematical details and intricacies, this course emphasizes the practical tools utilized to create DSP systems commonly used in real-world applications.

Description

This 3-day course provides an introduction to digital signal processing (DSP) tools and techniques used to analyze signals and design systems in one- and two-dimensions. DSP operations such as spectral estimation, frequency selective filtering, and image reconstruction from projections are treated. In contrast to typical DSP courses that focus on mathematical details and intricacies, this course emphasizes the practical tools utilized to create DSP systems commonly used in real-world applications.

MATLAB is used throughout the course to illustrate important DSP concepts and properties, permitting the attendees to develop an intuitive understanding of common DSP functions and operations. MATLAB routines are used to compute and interpret a signal’s frequency content, design and implement DSP filters for frequency selection, and reconstruct an unknown two-dimensional object from measured projection data.

The course is valuable to engineers and scientists who are entering the signal processing field or as a review for professionals who desire a cohesive overview of DSP with illustrations and applications using MATLAB. A comprehensive set of notes and references as well as all custom MATLAB routines used in the course will be provided to the attendees.

Course Outline:

1. Discrete-Time Signals & Systems.Frequency concepts in continuous- and discrete-time. Fourier Series and Fourier Transforms. Linear time-invariant systems, convolution, and filter frequency response
2. Sampling. The Sampling Theorem, Aliasing, and Signal Reconstruction.
3. The Discrete Fourier Transform (DFT) and Spectral Analysis. Definition and properties of the DFT, illustrated in MATLAB. Zero-padding, windowing, and efficient computational algorithms - the Fast Fourier Transform (FFTs). Circular Convolution, Linear Filtering with the FFT, and Overlap-save techniques. 
4. Design of Finite-Impulse Response (FIR) Filters. Filter Specifications in Magnitude and Phase. Requirements for linear phase. FIR filter design in MATLAB with the Windowing and Optimum Equiripple techniques.
5. Two-dimensional Signals Analysis.Frequency domain signal analysis concepts in 2-D. Sampling geometries in 2-D. 1-D to 2-D FIR filter conversion.
6. Computed Imaging Computed tomography, projections, and related techniques. The Radon transform and the Projection Slice Theorem. Image reconstruction via Direct Fourier Domain and Filtered Back-projection techniques.

What You Will Learn:

• Compute and interpret the frequency-domain content of a 1-D and a 2-D signal.
• Design and implement finite-impulse response (FIR) filters to satisfy a given set of specifications.
• Investigate computed tomography and related techniques to reconstruct an unknown object from projection measurements.
• Utilize MATLAB to analyze 1-D and 2-D digital signals, design and implement digital filters, and reconstruct an object from its projections.

Instructor(s):

Dr. Brian Jennison retired as a Principal Staff Engineer at the Johns Hopkins University Applied Physics Laboratory, where he worked on signal processing efforts for radar, sonar, chemical detectors, and other sensor systems. He holds M.S. and Ph.D. degrees in Electrical Engineering from Purdue University and a B.S. degree in Electrical Engineering from the Missouri University of Science and Technology. He served as Chair of the Electrical and Computer Engineering program for the Johns Hopkins University Engineering for Professionals, where he has taught courses in signals and systems, multi-dimensional and multi-rate digital signal processing.