A follow-up web page on Embedded
Computer Vision was created. And a few snapshots from
the tutorial are at the bottom of this page.
Overview
This tutorial covers a topic that is attracting more and more
interest: performing image processing and computer vision on
dedicated hardware, including right on the camera chip. The
tutorial is aimed at researchers, graduate students, and
practicioners of computer vision and will provide an overview and
introduction to hardware and software tools. We will discuss
algorithmic and hardware-specific considerations when
"outsourcing" computation to a DSP, FPGA, or other embedded
platform, and we will provide guidelines for how to best improve
the runtime performance of computer vision applications.
The tutorial will be held in conjunction with IEEE CVPR 2006 on
17 June 2006 at NYU in New York City. Once you have registered
for CVPR, please send us a quick email to embeddedCV@nps.edu with
your name - this will reserve a set of tutorial materials for you
(walk-ins are welcome, but we will have a limited amount of
readers/printed notes).
todo (say what CVPR registration option to use if participants
want to sign up for tutorial only)
When: afternoon session (1-5pm) on 17 June 2006
Where:Kimmel
Center at NYU, New York City
Who: for researchers, graduate students, and
practicioners of computer vision.
Outline/Syllabus
(this outline is tentative)
CPU or embedded? advantages and limitations
an example: automotive applications
hardware (DSP, FPGA, media chips/extensions, phones, smart
cameras, framegrabbers)
software and hardware tools for programming embedded
processors
