Signal Processing for Intelligent Sensor Systems with MATLAB

David C. Swanson

• Publisher: CRC Press
• Year: 2011
• eBook ISBN: 9781439896280
• Topic: Computer Science
• Subtopic: Computer Engineering
• Index: Computer Science

Signal Processing for Intelligent Sensors with MATLAB, Second Edition once again presents the key topics and salient information required for sensor design and application. Organized to make it accessible to engineers in school as well as those practicing in the field, this reference explores a broad array of subjects and is divided into sections:

Part I

Fundamentals of Digital Signal Processing

It was in the late 1970s that the author first learned about digital signal processing as a freshman electrical engineering student. Digital signals were a new technology and generally only existed inside computer programs and as hard disk files on cutting edge engineering projects. At the time, and reflected in the texts of that time, much of the emphasis was on the mathematics of a sampled signal, and how sampling made the signal different from the analog signal equivalent. Analog signal processing is very much a domain of applied mathematics, and looking back over 40 years later, it is quite remarkable how the equations we process easily today in a computer program were implemented eloquently in analog electronic circuits. Today there is little controversy about the equivalence of digital and analog signals except perhaps among audio extremists/purists. Our emphasis in this part is on explaining how signals are sampled, compressed, and reconstructed, how to filter signals, how to process signals creatively for images and audio, and how to process signal information “states” for engineering applications. We present how to manage the nonlinearity of converting a system defined mathematically in the analog s-plane to an equivalent system in the digital z-plane. These nonlinearities become small in a given low-frequency range as one increases the digital sample rate of the digital system, but numerical errors can become a problem if too much oversampling is done. There are also options for warping the frequency scale between digital and analog systems.

We present some interesting and useful applications of signal processing in the areas of audio signal processing, image processing, and tracking filters. This provides for a first semester course to cover the basics of digital signals and provide useful applications in audio and images in addition to the concept of signal kinematic states that are used to estimate and control the dynamics of a signal or system. Together these applications cover most of the signal processing people encounter in everyday life. This should help make the material interesting and accessible to students new to the field while avoiding too much theory and detailed mathematics. For example, we show frequency response functions for digital filters in this part, but we do not go into spectral processing of signals until Part II. This also allows some time for MATLAB^® use to develop where students can get used to making m-scripts and plots of simple functions. The application of fixed-gain tracking filters on a rocket launch example will make detailed use of signal state estimation and prediction as well as computer graphics in plotting multiple functions correctly. Also, using a digital photograph and two-dimensional low- and high-pass filters provide an interesting introduction to image processing using simple digital filters. Over 40 years ago, one could not imagine teaching signal processing fundamentals while covering such a broad range of applications. However, any cell phone today has all of these applications built in, such as sampling, filtering, and compression of the audio signal, image capture and filtering, and even a global positioning system (GPS) for estimating location, speed, and direction.

1 Sampled Data Systems

Figure 1.1 shows a basic general architecture that can be seen to depict most adaptive signal processing systems. The number of inputs to the system can be very large, especially for image processing sensor systems. Since an adaptive signal processing system is constructed using a computer, the inputs generally fall into the categories of analog “sensor” inputs from the physical world and digital inputs from other computers or human communication. The outputs also can be categorized into digital information, such as identified patterns, and analog outputs that may drive actuators (active electrical, mechanical, and/or acoustical sources) to instigate physical control over some part of the outside world. In this chapter, we examine the basic constructs of signal input, processing using digital filters, and output. While these very basic operations may seem rather simple compared to the algorithms presented later in the text, careful consideration is needed to insure a high-fidelity adaptive processing system. Figure 1.1 also shows how the adaptive processing can extract the salient information from the signal and automatically arrange it into XML (eXtensible Markup Language) databases, which allo...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Dedication
• Table of Contents
• Preface
• Acknowledgments
• Author
• Part I Fundamentals of Digital Signal Processing
• Part II Frequency Domain Processing
• Part III Adaptive System Identification and Filtering
• Part IV Wavenumber Sensor Systems
• Part V Signal Processing Applications
• Index