IMAGE PROCESSING AND MACHINE VISION

SHE Level 5
SCQF Credit Points 15.00
ECTS Credit Points 7.50
Module Code MMH623545
Module Leader Mario Mata
School School of Computing, Engineering and Built Environment
Subject Electronic Engineering
Trimester
  • B (January start)

Pre-Requisite Knowledge

Digital Programmable Systems 2 or equivalent Digital Signal Processing

Summary of Content

This module will introduce fundamental technologies for digital image/video representation and machine vision. This will introduce compression, analysis, and processing within vision systems. Students will gain understanding of algorithm, imaging system design, analytical tools, and practical implementations of various digital image algorithms and their subsequent application to real world machine vision problems. Image Acquisition. Image Enhancement, Compression and Segmentation. Image Processing. Multiresolution methods, Colour Processing, Object Recognition, Machine Vision and suitable Machine Learning techniques.

Syllabus

The teaching syllabus will cover the following areas: Image Acquisition: CCD/CMOS acquisition, Frame Grabbers, Pixel characteristics, Image resolution, image resolution. Image acquisition with area or line scan camera, Bayer grid. Image Enhancement and Filtering in the Spatial Domain: Histograms, Look-up Table (LUT) transformations, spatial filtering, smoothing filters, linear and non-linear filters, Gray level transformations, enhancement using arithmetic/logic functions. Image Filtering in the Frequency Domain: Fourier transform and the frequency domain, smoothing and sharpening frequency domain filters. Colour Spaces: The RGB space; alternative colour spaces, Hue Saturation Value(HSV), Hue Saturation Intensity (HIS), Hue Saturation Lightness (HSL) and conversions. Image Processing Image compression; image segmentation, detection of discontinuities, boundary detection, region-based segmentation, the use of motion in segmentation; Morphological Image processing, dilation, erosion, basic morphological algorithms, thresholding, binary hit-or-miss transformation. Face/Object Recognition and Tracking Background Subtraction. Colour based tracking. Viola-Jones face detection. Eigenfaces. Machine Vision Machine vision approach, effects of illumination, interface of machine vision components, image representation. Machine Learning Machine learning technique for classification within a machine vision system. Bayesian Methods, Ensemble Learning, Support Vector Machines. Image Processing Implementation Image processing algorithm implementation in Matlab and low level implementation in C within an embedded system.

Learning Outcomes

On completion of this module the student should be able to:Describe the characteristics of a range of algorithms used in image processing and relate these to typical applications in the field of Machine Vision.Develop software to manipulate images at the pixel level for colour, point, geometric and morphology transformationsCritically analyse and describe the various building blocks in a Machine Vision SystemUse industry standard software tools to assist with Digital Image Processing tasks.Use industry standard software tools to utilise Machine Learning within a Machine Vision system.

Teaching / Learning Strategy

The University 'Strategy for Learning' documentation has informed the learning and teaching strategy for this module. The course material will be introduced through lectures and practical exercises based on lecture material will be applied during lab and tutorial sessions. Tutorials will be used to explain and elaborate on the lecture material. Case studies will be used to demonstrate applications of image processing technology. A workshop will be given on the ethical aspects of machine vision and machine learning. The laboratory work will provide the student with support to develop both high level, and low level image processing implementation skills. This is achieved via a series of formative exercises, then a summative project hand in. To encourage the students to be as technically innovative as possible marks are given for both technical competence and creativity. Full use will be made of GCU Learn to provide Lecture-based and related study materials, along with sample solutions of Tutorial and Laboratory exercises, thus encouraging the development of independent learning and allowing self-reflective feedback on student performance. Staff-based feedback on student performance for submitted work will be provided in line with the University feedback policy, with summative feedback and grades on the coursework assessment utilising GCU Learn. The additional interactive discussion features of GCU Learn will be utilised, as appropriate to the module, to stimulate independent and flexible student learning outwith scheduled class time.

Indicative Reading

Both the learned and the popular literature on image processing and machine vision will be used as a source of information for private study. The titles and electronic resources below are to be considered as indicative only. -360 1. Digital image processing. Pearson International Edition 3rd Edition Rafael Gonzalez 2. Fundamentals of Digital Image Processing: A Practical Approach with Examples in Matlab, Wiley-Blackwell, Chris Soloman and Toby Breckon

Transferrable Skills

Specialist knowledge and application. Critical thinking and problem solving. Critical analysis. Communication skills, written, oral and listening. Numeracy. Effective Information retrieval and research skills. Computer literacy. Self confidence, self discipline & self reliance (independent working). Creativity, innovation & independent thinking. Knowledge of international affairs. Reliability, integrity, honesty and ethical awareness Presentation skills. Commercial awareness

Module Structure

Activity Total Hours
Lectures (FT) 12.00
Practicals (FT) 24.00
Tutorials (PT) 6.00
Practicals (PT) 12.00
Assessment (PT) 15.00
Independent Learning (PT) 105.00
Assessment (FT) 15.00
Independent Learning (FT) 93.00
Lectures (PT) 12.00
Tutorials (FT) 6.00

Assessment Methods

Component Duration Weighting Threshold Description
Exam (Exams Office) 2.00 50.00 45% Exam linked to Learning Outcomes
Coursework 1 n/a 25.00 45% Practical
Coursework 2 n/a 25.00 45% Practical