Title page for etd-0808117-123522


[Back to Results | New Search]

URN etd-0808117-123522
Author Chen-Chia Chung
Author's Email Address No Public.
Statistics This thesis had been viewed 5559 times. Download 0 times.
Department Electrical Engineering
Year 2017
Semester 1
Degree Master
Type of Document
Language zh-TW.Big5 Chinese
Title Error-Tolerability Enhancement for Videos by Adaptive I-Frame Insertion
Date of Defense 2017-09-07
Page Count 69
Keyword
  • quality improvement
  • video quality assessment
  • error-tolerability
  • erroneous video
  • video repair
  • Abstract With the advances in semiconductor technology, Internet of Things (IoT) devices and applications have been developing rapidly. Among these devices and applications, surveillance systems play an important role in environmental security and also lead to more convenient life. So far surveillance systems have become more and more popular, which makes enhancing the reliability of the associated video processing circuits more critical. Due to aging effects or external disturbances, videos may become erroneous and thus even unrecognizable. On the other hand, since surveillance systems usually have a large number of video sources, manual checking and repairing each video source by the maintenance staff may incur huge cost. Fortunately, errors in video processing circuits may not necessarily invalidate the entire surveillance system. Minor errors may not affect the success of identifying objects of interests in the video. As a result, how to efficiently assess the reliability of video sources is a critical issue to be addressed. Moreover, if the reliability can be further enhanced, the lifetime of video processing circuits can be effectively extended.
    In this thesis, we carry out detailed error simulations for a video decoder, and analyze the impact of errors on the quality of videos. We also investigate the problems that recent video repair methods would encounter in repairing erroneous videos resulting from faulty decoders. Accordingly we develop an effective video quality enhancement technique and identify the scenarios that the developed technique is applicable. In these scenarios at least 77% of unacceptable erroneous videos can become acceptable.
    In addition, we also propose an error-tolerability enhancement flow for videos. In particular, this flow integrates with the no-reference error-tolerability test method developed by our research group. This integration leads to 90% test time and 99% needed memory space reduction when compared with the previous video quality assessment method. In addition to testing the error-tolerability of the circuit, our proposed flow can also further improve the error-tolerability of the circuit.
    Advisory Committee
  • Shiann-Rong Kuang - chair
  • Hsin-Wen Ting - co-chair
  • Jih-Ching Chiu - co-chair
  • Tong-Yu Hsieh - advisor
  • Files
  • etd-0808117-123522.pdf
  • Indicate in-campus at 5 year and off-campus access at 5 year.
    Date of Submission 2017-09-11

    [Back to Results | New Search]


    Browse | Search All Available ETDs

    If you have more questions or technical problems, please contact eThesys