Machine Sensing

Module summary

Module code: ELEE1161
Level: 7
Credits: 15
School: Engineering and Science
Department: Engineering
Module Coordinator(s): Seyed Pedram

Specification

Aims

• To develop an in-depth understanding of digital signal processing (DSP) & machine sensing (audio, listening, video, vision), associated numerical analysis and its application to engineering systems.
• To provide theoretical and practical skills in the establishment of a variety of computer modelling and simulation techniques.
• To provide students with advanced skills in the use of mathematical simulation packages and hardware devices.

Learning outcomes

On successful completion of this course a student will be able to:

1 Reflectively compare and contrast various available technologies for data acquisition hardware with particular application in machine sensing.

2 Demonstrate an in-depth understanding of the principles and techniques of digital representation of audio and video signals with particular application in machine sensing.

3 Formulate innovative audio and image digital processing solutions using commercial software and hardware with application in machine sensing.


Indicative content

The focus of this module is the application of digital signal processing techniques (DSP) to machine sensing systems. Essentially, machine sensing is the ability of a computer to ‘see’ and ‘hear’ their surroundings just a human being would by using electronic systems that perceive and ‘understand’ electronic signals. For example, machine sensing can be used to determine manufacturing faults within components, facial/voice recognition systems, and even the ability to ‘see’ inside the human body and other objects. There are two distinct processes involved in machine sensing systems, these processes being data acquisition from hardware devices and data processing using electronic and software applications; this module has been designed to provide skill and knowledge in both processes.

The module will typically cover subjects such as applications of machine sensing applications, data acquisition hardware devices, analogue-to-digital conversion (ADC), discrete systems, sampling mechanisms, linearity, image types (classification, formation, geometry), digitisation, pixels, quantisation, digital correlation, neighbourhood construction, sources of noise, noise removal, image pre-processing, data compression, digital filters, adaptive filtering, thresholding, image enhancement, Fourier transforms, image restoration, morphological processes, image segmentation, audio standards, speech processing (coding, recognition), video restoration, recursive/non-recursive audio filters, sample rate conversion, digital audio interfaces, machine learning, sensing and motion, optical process tomography (OPT), and graphical user interface (GUI) design.

Teaching and learning activity

The module will be taught via weekly 2hr lectures and 2hr computer modelling laboratories. Electronic copies of lecture notes will be made available for download each week throughout the module. The coursework element will comprise of a group based computer modelling challenge. Typical challenges will involve developing image processing software to analyse and remove noise from images, to develop software to detect abnormalities within artefacts, or to develop software to facilitate audio acquisition and enhancement. An industry leading simulator (Matlab) will be extensively used throughout the module to establish computer models for analysis and to facilitate comparison of numerical and simulated data. Formative online quizzes will be used to identify student progress and provide detailed feedback.

Assessment

Students are required to pass all components in order to pass the course.

Methods of SUMMATIVE Assessment: Report
Outcome(s) assessed by summative assessment: 1-3
Grading Mode: Numeric.
Weighting: 50%.
Pass Mark: 50%
Work Length: 7 pages.
Outline Details: Computer challenge, Group report (2-pages), Individual report (5-pages)

Methods of SUMMATIVE Assessment: Examination.
Outcome(s) assessed by summative assessment: 1 - 2
Grading Mode: Numeric.
Weighting: 50%.
Pass Mark: 50%.
Outline Details:Comprising of a 2hr closed book end of year examination (90%) and a 1 hour online examination (10%)

Nature of FORMATIVE assessment supporting student learning:
Online quizzes and examination practice tests held throughout the module.