Talks To TR100 Winner Ramesh Rasker
(This article is sponsored by Attorney
Dr. Ramesh Raskar, a research scientist at
Mitsubishi Electric Research Lab in Cambridge, MA, is one of the nine
Indians who have won the prestigious TR100 award given by the MIT
Technology Review magazine. The chosen 100 (TR100) represent a group under
the age of 35 that is using technology to transform the world around us.
According to the Technology Review magazine Raskar was chosen for his work
on building large computer display systems that seamlessly combine images
from multiple projectors. This work may lead to new applications in
entertainment, image-guided surgery, and user interfaces.
Ramesh Raskar joined MERL as a Research Scientist
in 2000 after his doctoral research at U. of North Carolina at Chapel
Hill, where he developed a framework for projector based displays. His
work spans a range of topics in computer vision and graphics including
projective geometry, non-photorealistic rendering and intelligent user
interfaces. He has developed algorithms for image projection on planar,
non-planar and quadric curved surfaces that simplify constraints on
conventional displays and has proposed Shader Lamps, a new approach for
projector-based augmented reality. Current projects include composite RFID
(RFIG), multi-flash non-photorealistic camera for depth edge detection,
locale-aware mobile projectors, high dynamic range video, image fusion for
context enhancement and quadric transfer methods for multi-projector
curved screen displays.
Dr. Raskar has also received the Mitsubishi
Electric Information Technology R&D Award 2003, Global Indus
Technovator Award 2003, instituted at MIT to recognize the top 20 Indian
technology innovators on the globe and the Mitsubishi Electric Valuable
Invention Award 2004 His papers have appeared in SIGGRAPH, Eurographics,
IEEE Visualization, CVPR and many other graphics and vision conferences.
He has taught courses and has served as a member of international program
committees at major conferences. He is a member of the ACM and
Dr. Raskar took us on a tour of his lab at MERL and
gave us some insights into his work and life.
Lokvani: What problem does your work on
seamlessly combining images from multiple projectors try to address?
Raskar: The motivation for the work was a desire to replicate the
shift that happened in the world of computing in the world of displays.
Originally supercomputers were one large unit, expensive to build and
difficult to maintain. Today supercomputers are built by weaving together
off-the shelf Linux boxes.
Large computer displays, like the kind
found in digital cinema, IMAX theaters, planetariums or scientific
visualization centers are commonly created using a single expensive high
powered projectors. Such a system is not easy to maintain since a failure
in one of the projectors would bring the entire system down. These systems
are not very flexible and cannot be upgraded easily.
large computer displays by automatically stitching together images that
are beamed by off-the shelf projectors. This system is cheaper to build
and is very flexible. We can recreate a good quality image rather easily
even if a projector fails which makes fault tolerant. Plus individual
units can be upgraded anytime allowing the owner to ride the technology
and cost curve.
Lokvani: Does this technology have
applications beyond the entertainment industry?
technology has a variety of applications. In the future, every light bulb
could be replaced with a projector at home and at the work place. This
could enable us to beam a variety of images on different surfaces. When
you travel for work, you may be able to have live teleconferencing with
different members of your family and thus be able to create a virtual
reality “at-home” feel.
Lokvani: Could you tell us a little about
your other innovations viz the multi-flash camera for shape boundary
detection and RFIG?
Raskar: Imagine a camera, no larger than
existing digital cameras, that can directly create drawings that you see
in car manuals or anatomy books. As we know, a flash to the left of a
camera creates a sliver of shadow to the right of each silhouette in the
image. We add a flash on the right, which creates a sliver of shadow to
the left of each silhouette, a flash to the top and bottom. Simply by
observing the shadows, after image processing, we can robustly find all
the pixels corresponding to shape boundaries. This is a strikingly simple
way of calculating the so called depth edges and solves a fundamental
problem in Computer Vision. Since we can distinguish between color
boundaries and shape boundaries, we can use these edges to create images
that look like technical illustrations and line drawings. We have a built
a new type of laparoscope by adding multiple light fibers around the tiny
camera of the laparoscope. Inside the patient, the lit fibers cast shadows
around lesions or minute shapes in different directions and after image
processing, we can provide live images during examination or surgery that
are much easier to understand.
RFIG is a photosensing wireless tag
which acts as a radio frequency identity and geometry transponder. Think
of it as a millimeter-accurate location aware RFID tag. We find the
location of the photosensing RFID tag using a handheld ‘projector’ which
looks like your TV remote control that beams infrared light. We have
introduced a novel technique that we call interactive projection to allow
a user to interact with projected information e.g. to navigate or update
the projected information.
Lokvani: You have been a successful
innovator and won several prestigious awards. To what do you attribute you
Raskar: The work environment. MERL allows me to take my
ideas and provides me all the resources to work on it. The flexible
structure of the organization allows me to draw upon people who have
different expertise and work in different areas. MERL allows us to
participate in any aspect of innovation whether it is doing fundamental
research, acquiring patents or transform your idea into a business
opportunity. This has greatly contributed to my success.
I am very
grateful for all the guidance and support I have received from my advisor
at the University of North Carolina, Dr. Henry Fuchs and Dr. Greg Welch
and the director of MERL, Dr. Joe Marks.
I am most grateful to my
older brother, Prakash, who has always been there to provide a sense of
balance. He picks me up when I am down and helps me understand that awards
are really nothing more than a small token of appreciation for my work.
Lokvani: You have included the Sanskrit
verse in your web page that gives the value of pi to ten decimal places.
Do you have a special interest in the area of Vedic
Raskar: People in the US often ask me why Indians are
so mathematically inclined. This made me investigate our history a little
to understand where this interest comes from. This search led me to this
verse and others which I find fascinating.
Ever since I put this
verse on my web page I have received calls from many interesting people
including a film producer who is making a film on the origins of pi.
Lokvani: What are your plans for the future?
Raskar: On the professional front I hope to continue to work on
the different technologies that we have developed. Venture Capitalists
have shown interest in some of the technologies.
On the personal
front, I am still single, but too busy and hope I can find the right
Lokvani: On behalf of Lokvani, I wish you
the very best of luck in finding a soul mate. We thank you for your
Raskar: Thank you.
To learn more please visit Dr. Raskar's website at
Article | Share
You may also access this article through our web-site http://www.lokvani.com/