ECE352 Image Processing

Topics covered include 2-D signal sampling and quantization, image storage and retrieval in popular formats, image enhancement, intensity transformations, filtering in the spatial and frequency domains, 2-D Fast Fourier Transform, interpolation and image resampling, geometric image manipulation, texture identification and segmentation, color image processing, morphological image processing, halftoning, image restoration and image reconstruction from projections, and image compression. All topics are covered in the classroom with a presentation of corresponding theory and examples, followed by C / C++ / Python language implementations, optimization techniques and finally by homework assignments that implement some of the topics. A final project is required by each student that gives the opportunity to apply both analytical and synthesis skills, starting with their ability to analyze and provide solution to an image processing task, implement it efficiently in software, and present it appropriately through a project report.

In titles, subject covered are:

• 2-D signal sampling and quantization
• Image storage and retrieval in popular formats
• Image enhancement– intensity transformations
• Filtering in the spatial and frequency domains
• 2-D Fast Fourier Transform
• Interpolation and image resampling
• Geometric image manipulation
• Texture identification and segmentation
• Color image representation and processing
• Morphological image processing
• Halftoning
• Image restoration and image reconstruction from projections
• Image compression

Homework assignments and final project are also significant components of the course.

This course introduces students to digital image processing fundamentals, algorithms and applications.

It is an advanced, application-oriented class that provides students with opportunities to develop their own ideas into useful applications, to learn professional software development tools and techniques, and test their knowledge against research topics. It can further provide the starting point for advanced topics in computer vision that can be coordinated at later semesters and finally the basis for diploma theses.

By the end of the course, students must be able to manipulate digital images in a variety of ways, process images with their own software, work with professional software development tools in personal computers, find and fix logical problems in complex algorithms, optimize time-critical functions and explain results in scientific/engineering terms.

Typical students will have acquired the following skills:

• Understanding the nature, characteristics of digital images and applications of general signal-processing techniques on images
• Ability to select the appropriate algorithms to perform various tasks on digital images
• Develop high-quality C / C++ / Python language software that implements image processing tasks
• Present and explain their results in engineering terms and a professional manner
• The ability to design their own algorithms by combining the knowledge acquired in order to solve new problems
• The ability to evaluate the results of their work, by comparing them with those provided by the theory and often with those resulting from alternative methods
• The skills to continue studying more advanced related subjects