Reducing matching time for OPS5 production systems

Kang, Jaewoo; Cheng, Albert M. K.

doi:10.1109/cmpsac.2001.960649

Cited by 3 publications

(1 citation statement)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Currently, important parts of a given image are identified by a human user. Our on-going work employs machine learning [27,28] to allow automatic identification of the important parts in an image within specific applications domains.…”

Section: Discussionmentioning

confidence: 99%

Priority-driven Coding and Transmission of Progressive JPEG Images for Real-Time Applications

Cheng

Feng

2007

J VLSI Sign Process Syst Sign Image Video Technol

View full text Add to dashboard Cite

Since high-quality image/video systems based on the JPEG/MPEG compression standards often require power-expensive implementations at relatively high bit-rates, they have not been widely used in lowpower wireless applications. To alleviate this problem, we designed, implemented, and evaluated a strategy that can adapt to different compression and transmission rates. (1) It gives important parts of an image higher priority over unimportant parts. Therefore, the high-priority parts can achieve high image quality, while the lowpriority parts, with a slight sacrifice of quality, can achieve huge compression rate and thus save the power/ energy of a low-power wireless system. (2) We also introduce a priority-driven scheduling approach into our coding algorithm, which makes the transmission of important parts earlier with more data than other parts. Through a balanced trade-off between the available time/bandwidth/power and the image quality, this adaptive strategy can satisfy users with desired images quality and lead to a significant reduction of the important parts_ deadline misses.

show abstract

Section: Discussionmentioning

confidence: 99%