Grant: Discovery Grants Program - Individual
University of Saskatchewan
$19,491.00
New architectures for real-time video and bio-medical imaging applications
EF 2012-13
Résumé
Most real-time multimedia-based applications employ digital video and image processors that are built on computation of irrational transform coefficients (such as cosine, sine, or wavelet functions), which cannot be exactly implemented due to lack of finite binary representation. Hence, in implementing them, we invariably introduce heavy truncation errors; as the process continues, more errors are incurred, which eventually results in inefficient design, and degrades reconstruction quality. In the recent past, error-free scheme has been shown to be an effective way in eliminating such truncation errors, where the basis functions are exactly encoded with integers. This integer implementation is very simple, and enables building efficient VLSI architectures.
The research expands and refines the concept of ""error-free computation"" by first proposing an investigation in a novel multi-dimensional level that allows more implementation-friendly features (such as parallelism, pipelining, simple additions, etc.), resulting in low-area, low-power, and high-speed designs. Secondly, by taking advantage of these attributes, we aim to build efficient and performance-enhanced intellectual property cores such as: (a) transform units for JPEG, JPEG2000, and H.264 compliant encoders/decoders with high quality data reconstruction for real-time applications, (b) video compressor for biomedical imaging applications such as capsule endoscope and telemedicine, and (c) image compressor for information security applications such as digital watermarking. These architectures will be implemented on both field programmable gate arrays (FPGA) and standard-cell platforms so that the performance improvements in hardware and application levels can be demonstrated. Finally, we aim to develop a novel ""generalized error-free encoding"" that will encode virtually any discrete transform coefficient error-freely so that other designers may benefit from this scheme. The results of this research will significantly enhance the efficiency and the accuracy of many video-related commercial applications, which can be transferred to Canadian industry, enhancing Canada's strong role in defining future video products.
Programme
Discovery Grants Program - Individual
Bénéficiaire
Emplacement
Saskatchewan, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data
Tel que publié
Application id
352535-2008
Name nom
Wahid, Khan
Department département
Electrical & Computer Engineering
Organization id
16
Institution établissement
University of Saskatchewan
Province en
Saskatchewan
Province fr
Saskatchewan
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2012
Competition year année de concours
2008
Award amount
19491
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
Discovery grants programs
Group fr
Programmes de subventions à la découverte
Committee code
334
Committee name en
Commun.,Computers & Components
Committee name fr
Communic.,ordin. et composants
Area of application code
802
Area of application group en
Information and communication services
Area of application group fr
Services d'information et de télécommunications
Area of application en
Information systems and technology
Area of application fr
Systèmes d'information et technologie de l'information
Research subject code
2708
Research subject group en
Information technology
Research subject group fr
Technologie de l'information
Research subject en
Image processing
Research subject fr
Traitement des images
Partie
5
Nb partie
5
Application title
New architectures for real-time video and bio-medical imaging applications
Keyword
NSERC
Application summary
Most real-time multimedia-based applications employ digital video and image processors that are built on computation of irrational transform coefficients (such as cosine, sine, or wavelet functions), which cannot be exactly implemented due to lack of finite binary representation. Hence, in implementing them, we invariably introduce heavy truncation errors; as the process continues, more errors are incurred, which eventually results in inefficient design, and degrades reconstruction quality. In the recent past, error-free scheme has been shown to be an effective way in eliminating such truncation errors, where the basis functions are exactly encoded with integers. This integer implementation is very simple, and enables building efficient VLSI architectures.
The research expands and refines the concept of ""error-free computation"" by first proposing an investigation in a novel multi-dimensional level that allows more implementation-friendly features (such as parallelism, pipelining, simple additions, etc.), resulting in low-area, low-power, and high-speed designs. Secondly, by taking advantage of these attributes, we aim to build efficient and performance-enhanced intellectual property cores such as: (a) transform units for JPEG, JPEG2000, and H.264 compliant encoders/decoders with high quality data reconstruction for real-time applications, (b) video compressor for biomedical imaging applications such as capsule endoscope and telemedicine, and (c) image compressor for information security applications such as digital watermarking. These architectures will be implemented on both field programmable gate arrays (FPGA) and standard-cell platforms so that the performance improvements in hardware and application levels can be demonstrated. Finally, we aim to develop a novel ""generalized error-free encoding"" that will encode virtually any discrete transform coefficient error-freely so that other designers may benefit from this scheme. The results of this research will significantly enhance the efficiency and the accuracy of many video-related commercial applications, which can be transferred to Canadian industry, enhancing Canada's strong role in defining future video products.