|
|
 |
News
Archives
Recent
AMA,
Inc. Awarded Mobile OCR Research Project
April 2006 - Applied Media Analysis, Inc was
recently awarded the Mobile OCR SBIR (Small Business Innovation
Research) Phase I project from the National Eye Institute, under
the Department of Health and Human Services. For this project
AMA will demonstrate the technical feasibility of Mobile OCR,
a portable software module which makes use of existing personal
devices to provide access to textual materials for the elderly
or the visually impaired. The system will help these low vision
individuals with basic activities in their daily lives, such
as shopping, preparing meals, taking medication, and reading
traffic signs. It will step beyond AMA's previous MobileEyes
vision enhancement system to apply cutting edge recognition technology
for mobile devices. The system will use common camera phone hardware
to capture and enhance textual information, perform Optical Character
Recognition (OCR) and provide feedback through audio or visual
means.
AMA,
Inc. Awarded NIH Research Project
September 2005 - Applied Media Analysis,
Inc. was awarded a Small Business Innovation Research (SBIR) project
to provide a software based vision enhancer for the visually impaired.
Given the importance and high cost of healthcare for
the visually impaired, particularly elderly adults, applications
of technology that provide visual enhancement
are especially
valuable.
For this project, AMA demonstrated the technical feasibility of
MobileEyes, a software based add-on suite which uses existing camera
enabled PDA or cell phone hardware to provide enhanced electronic
imaging capabilities. The
system provides video based magnification and contrast enhancement,
with specialized enhancement for text making it easily read by low
vision users. The system is also configurable to account for
the specific needs of the given user.
The technical
contributions of this project focused on text enhancement since
an inability to read directly impacts daily activities such as
shopping,
preparing
meals, taking medication, or even using household appliances.
AMA,
Inc. Awarded IBARS Project Phase II
September 2005- After successfully completing IBARS Project Phase I,
Applied Media Analysis, Inc (AMA) was awarded Phase II. Based on
the previous
concept of using available mobile devices to link the physical
world to information networks, AMA’s proposed mobile symbol recognition
technology from Phase I enabled many opportunities for mobile
e-commerce by recognizing bar codes, text on documents, and user-customizable
icons that carry and convey information. Our concept
remains powerful
in that it requires no new infrastructure, since it uses popular
mobile
devices, and existing symbols such as barcode tags, text, and user-customizable
icons. Once completed, our downloadable symbol recognition component
will enable many functions including sales, order-fulfillment,
targeted information-delivery, etc. For our customer and partner
base, service-providers, OEMs, merchants, advertisers,
and information providers are apt canidates.
Combined with
our Phase III partner, Nokia’s Mobile Commerce Solutions, our
cross-platform symbol recognition technology is well positioned
to capture a nontrivial
portion of
this large mobile market, and make significant contributions
to the US economy.
Finally, we are also seeking to use our technology to help
disadvantaged groups (handicapped or visually impaired, for example)
get access
to product information (prescription drug instructions, for
example) or
transact commerce activity conveniently, using a device they
may already have, or that is easily acquired. These include
applications in medical
care delivery, military applications, sign recognition for
the visually challenged, and others.
AMA,
Inc. Awarded IBARS Project
January 2004-Applied Media Analysis, Inc. (AMA) was
recently awarded an NSF SBIR (Small Business Innovation Research)
Phase I project.
This project sought to leverage the convergence of processing and
sensor technologies in widely available
camera-equipped cellular telephones in order to develop
e-commerce
applications centered on the acquisition and recognition of barcode
images. Modern
handheld devices present a convergence of many technologies in
a handy package,
including networking, voice, cameras, processing, location-sensing
and displays. However, small-size and long battery-life requirements
lead to
limited processing power, limited-resolution cameras, and varying
available network bandwidth on such handheld devices. To make
our concept feasible, improvements were needed in the algorithms
performing computer-vision,
image-processing, and pattern-recognition, so that they were
both computationally efficient
and small enough to fit in the memory of consumer devices. Challenges
AMA overcame included being able to unwarp images to account
for distortions due to perspective imaging, warping of the substrate,
or
non-flat
surfaces.
The image needed to be processed to account for imaging limitiations
such as non-uniform lighting, blurring, and highlights. The recognition
algorithms
in the system were created to be able to recognize many symbologies,
make use of extra information available in images to tackle degradations,
and be efficient
and small. Phase I research tackled these issues and demonstrated
project feasibility. The technology developed under this SBIR
initiative will facilitate the
development of numerous applications where by users can get and
communicate information through coded symbologies in many domains.
Specialized applications, such as the use of coding on advertisements,
direct
integration of mobile-information
and mobile-commerce, tracking of dietary restrictions, or product
contents from UPC codes,
are just
a few of the possible uses for this technology.
AMA,
Inc. Awarded MATES Project Phase II
November 2004-After successfully completing the MATEs Project Phase
I, Applied Media Analysis, Inc (AMA) was awarded Phase II. The Phase
I effort was comprised of three top-level objectives, all of which
have been completed and demonstrated key capabilities. For the
first objective, AMA
gathered data and developed,
demonstrated, and evaluated a set of algorithms for the detection,
recognition, and translation of road signs from images captured
by a PDA. Next, AMA
established a software architecture to support the processing,
including detection, recognition and translation of signs in various
domains, and
created an interface for configuring the system, providing feedback
to the user, visualizing results and integrating with other applications.
AMA implemented their analysis algorithms in all of these areas.
And
finally, AMA fostered existing collaboration, established new partnerships
and is developing a detailed plan for transitioning the technology
into the commercial sector.
Phase II of the project will build upon the successes of Phase I and
move toward full deployment of a downloadable software module for the
commercial market by the end of Phase II. AMA will continue to develop
a specific configuration and to address needs of the government sponsors
and the soldiers that will use the system. Phase II technical objectives
are: to realize the full technical development of our expanded system,
achieve commercial product usability, and develop a comprehensive technology
transition plan, fostering existing commercial interests, and promoting
new ones.
AMA,
Inc. Awarded MATES Project
January 2003 - Applied Media Analysis, Inc. has been
awarded a Phase I SBIR (Small Business Innovative Research) from
the Army Research Laboratory. Our mission was to develop
a sign understanding and translation system called MATES (Multilingual
Automatic Translation Engine for Signs) that can be operated by
a novice user to obtain and interpret signs in foreign languages.
The system, once completed, will
be comprised of a hand-held Personal Digital Assistant (PDA) and
camera along with an embedded component architecture and software
developed for text detection, recognition and translation, and
a configuration and visualization interface to aid personnel deployed
in foreign
countries.
Although the initial phase of this project will focus on the immediate
need for road sign translation, an architecture will be set up
to allow a range of domain knowledge modules to be loaded for translations
of
other types of signs, including posters, public announcements,
and documents left by adversaries, as well as more general text
content
found on menus,
in bus schedules, or with instructions.
Our text detection and recognition approach builds upon our previous
Research and Development to provide a lightweight, trainable, and
robust solution
which can be easily adapted to different foreign languages.
Our novelty lies on the use of context for both recognition and
translation/transliteration. In order to deal with uncertainty
in OCR and in the translation of sign
content, we rely heavily on contextual information and optional
feedback provided by the user of the system, combined with robust
techniques for indexing and incorporating domain information. The
general philosophy
will be to provide a system where additional context and resources
will be supported by dynamic configuration. The component architecture
supports
a view such that different resources and software components
can be loaded on demand.
The benefit of the development of an autonomous sign translation system
has tremendous commercial potential. For example: • Assisting visitors to foreign countries that are unable to read local language
• Sign detection and recognition can be integrated into systems for the visually impaired
• Sign recognitions module can be used in numerous applications including automated
mapping, robot navigation, and driver support systems
AMA,
Inc. teams with Perceptek Robotics of Littleton, Colorado
March 2002- Applied Media Analysis,
Inc. has entered into a partnership with Perceptek Robotics of
Littleton, Colorado
to
develop
sign
recognition capabilities for autonomous vehicles and driver assisted
systems. Under the support of a DARPA Phase I SBIR (Small Business
Innovative Research), we will build upon previous Research and
Development of both companies to provide a real-time architecture
for detecting, tracking, and recognizing signs from dynamic video
imagery. Our approach detects signs by combining visual cues consisting
of color, shape, and text. Signs are tracked to improve detection
accuracy and provide resiliency to occlusions and environmental
effects. The orientation of the sign is determined and the imagery
rectified for recognition processing. Text and graphics characters
are extracted using a combination of optical character recognition
and syntactic/semantic parsing. Feedback loops are provided to
resolve ambiguities, improve recognition rates, and reduce false
detections. Our Phase I program will result in thorough testing
of the algorithms and an end-to-end feasibility demonstration.
|
|
