Assignment 1
Topic: Background Segmentation
Due on or before:
13 February (Tuesday), 2024
Maximum Marks: 9
Here is a video [please click on the above image to play/save/download
the video] of two people loitering about aimlessly in a parking
lot in front of the A. V. Williams Building, University of
Maryland, College Park. (You may perhaps recognise the 6-footer
in a greyish-blue shirt and dark trousers: he was quite lean and
thin at that time...and healthy!)
Are you sick of him? Shoot him (with a camera)... Eliminate
him from the picture...
Can you perform background subtraction on this video?
Use the idea of motion segmentation, from the papers below. To segment
out the foreground, you may use the Sequential Labelling algorithm.
Or K-Means, for that matter.
Please generate two videos (in terms of the frames, and
the video file as well) -
-
the background, and
-
one with the foreground objects
Please also submit an ASCII .txt file, which contains a
description of the specific parameters used.
The recommended programming language is C/C++/Java/Python, and the
recommended software environment is
OpenCV. No MATLAB for this
assignment, please.
This assignment aims at getting experience with Machine
Learning algorithms which can easily be coded up. (There are many
which cannot, and may be used as black boxes). Even in OpenCV
(and other places), one may find numerous versions of the source
code: for Gaussian Mixtures, K-Means, and Stauffer and Grimson's
Background subtraction itself. Please do not copy/cut-and-paste
code from the Internet: this will be penalised, as mentioned on
the course webpage (the front page). You may read up different
implementations, but please code from scratch.
How does one handle MPEG videos?
Please feel free to use any video player to play MPEG video, and
any decoder/encoder of your choice, on an operating system of
your choice.
You can download free MPEG software from the MPEG website at
http://www.mpeg.org/
This contains links to many resources - free, and otherwise.
Of particular importance to you may be the UC Berkeley
MPEG player
mpeg_play,
and the MPEG Software Simulation Group's MPEG encoder and
decoder
mpeg2encode, and
mpeg2decode,
all of which are downloadable through the links - including the
source code!
Internal Links [IIT Delhi]
Demo Schedule:
Venue: (to be announced)
Demos:
(To be announced)
Those who stay outside the campus and may have some difficulty in
manitaining the above timings, are to contact the TAs
over email, well in advance. Ditto for others, who may have a
valid reason for not adhering to the above schedule.
Their demos will be rescheduled appropriately.
Sumantra Dutta Roy,
Seshan Srirangarajan,
Department of Electrical Engineering, IIT Delhi, Hauz Khas,
New Delhi - 110 016, INDIA.
sumantra@ee.iitd.ac.in
seshan@ee.iitd.ac.in