[ Home ] [ Syllabus ] [ Schedule ] [ Readings ] [ Assignments ] [ Links ]
Mobile Imaging - Winter Quarter 2014
This schedule will be updated occasionally...
PROJECT info here (link to come)
Assignment #1 - due 1:00pm on Thursday, January 9
- Consider imaging, and mobile imaging, topics that you would be interested in covering during the quarter. Come to class ready to offer them, perhaps briefly discuss your interest in them.
- Send an email to Emily with the information requested here.
- Send email to the instructor - one per group - with your first, second, and third choice of papers to present. You may choose from the papers here or choose your own. The paper should be a journal (or in some cases a top conference) research paper on a topic relevant to mobile imaging. If you're not sure where to look, you can peruse the list of "journals of interest" on the Links page.
- Read and understand your assigned paper well; this probably entails also reading some of the background material (concepts and research leading up to the paper). Go through it and discuss it together; form opinions. Strategize on how to best communicate the work to the class.
- Presentations: 15 minute presentation (+ 5 minute questions/discussion) by the team - everyone participate. Clearly present the work that was done, including the key problem(s) being addressed, the context (background) of the work, the approach, the results, how the outcomes were assessed, and your views on the paper. Don't just "go through slides" - make sure you communicate the key information to the audience. The class will grade each presentation based on how well they were able to communicate the research.
Assignment #3 - due on Tuesday, February 11
- The intent of this assignment is to get everyone started on mobile imaging app development with a simple example.
- Develop and demonstrate a mobile app with the following functionality:
- The live camera input is displayed at start-up, along with the GPS coordinates of the device and the accelerometer and gyroscope values.
- The live image may optionally be displayed as (a) grayscale, (b) binary (with an adjustable threshold), (c) intensity-normalized (where r+g+b is constant), or (d) gamma-corrected (with the ability to set the gamma parameter, a positive scalar. (Some of these may be done directly through OpenCV calls; others may require direct manipulation of pixel data.)
- When tapped, the image is saved to file in its current display type (RGB, grayscale, binary, or intensity-normalized) in JPG format. The filename should be image_N.jpg, where N is incremented from 0 for each successive saved image (including past runs of the app).
- Turn in your source code (email a zip file to the instructor)
- Demonstrate the app to the instructor
- For this assignment, you may ask anyone for help and advice, but you must develop your own app. Share ideas and know-how, but not code.
Project report and presentation - due on Tuesday, March 18
- Project information available here.