• Journal of information and communication convergence engineering
• /
• 제5권4호
• /
• pp.339-345
• /
• 2007

Video transcoding is a technique used to convert a compressed input video stream with an arbitrary format, size, and bitrate into a different attribute video stream different attributes to provide a efficient video streaming service for the customers is dispersed in the heterogeneous networks. Specifically, frames deletion occur in a transcoding scheme that exploits the adjustment of frame rate, and at this time, the loss in temporal relation among frames due to frame deletion is compensated for the prediction of motion estimation by reusing motion vectors in the would-be deleted frames. But the processing time for transcoding don't have an improvement as much as our expectation because transcoding is done only within the transcoder. So in this paper, we propose a new transcoding algorithm based on prediction period to improve transcoding-related processing time. For this, we also modify the existing encoder so as to adjust dynamically frame rate based on the prediction period and deletion period of frames. To check how the proposed algorithm works nicely, we implement a video streaming system with the new transcoder and encoder to which it is applied. The result of the performance test shows that the streaming system with proposed algorithm improve 60% above in processing time and also PSNR have a good performance while the quality of pictures is preserved.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제6권11호
• /
• pp.2827-2848
• /
• 2012

Previously, we described a social media cloud computing service environment (SMCCSE). This SMCCSE supports the development of social networking services (SNSs) that include audio, image, and video formats. A social media cloud computing PaaS platform, a core component in a SMCCSE, processes large amounts of social media in a parallel and distributed manner for supporting a reliable SNS. Here, we propose a Hadoop-based multimedia system for image and video transcoding processing, necessary functions of our PaaS platform. Our system consists of two modules, including an image transcoding module and a video transcoding module. We also design and implement the system by using a MapReduce framework running on a Hadoop Distributed File System (HDFS) and the media processing libraries Xuggler and JAI. In this way, our system exponentially reduces the encoding time for transcoding large amounts of image and video files into specific formats depending on user-requested options (such as resolution, bit rate, and frame rate). In order to evaluate system performance, we measure the total image and video transcoding time for image and video data sets, respectively, under various experimental conditions. In addition, we compare the video transcoding performance of our cloud-based approach with that of the traditional frame-level parallel processing-based approach. Based on experiments performed on a 28-node cluster, the proposed Hadoop-based multimedia transcoding system delivers excellent speed and quality.

• 산업기술연구
• /
• 제27권B호
• /
• pp.185-192
• /
• 2007

Transcoding video proxy is necessary to support various bandwidth requirements for mobile multimedia and to provide adapting video streams to mobile clients. Caching algorithms for proxy are to reduce the network traffic between the content servers and the proxy. This paper proposes a Meta-tailed caching for transcoding proxy that is efficient to lower network load and CPU load. Caching of two different data types - transcoded video, and metadata - provides a foundation to achieve superior balance between network resource and computation resource at transcoding proxies. Experimental results show that the Meta-tailed caching lowers at least 10% of CPU-load and at least 9% of network-load at a transcoding proxy.

• 정보처리학회논문지:컴퓨터 및 통신 시스템
• /
• 제3권7호
• /
• pp.231-240
• /
• 2014

스트리밍 서비스에 사용되는 트랜스코딩 서버는 동시에 다수의 트랜스코딩 요청을 처리한다. 하지만 트랜스코딩 연산의 부하가 높기 때문에 단일 서버가 수용할 수 있는 요청 개수는 제한적일 수밖에 없다. 본 논문에서는 단일 서버의 트랜스코딩 성능을 높이고자 CPU 기반 트랜스코더와 GPU 기반 트랜스코더를 동시에 사용하는 듀얼 트랜스코딩 방법을 제안한다. 듀얼 트랜스코딩 방법은 트랜스코딩 요청을 처리하기 전에 워크로드를 예측하여 해당 요청에 대한 QoS유지가 가능한지 판단한다. QoS유지가 가능하다고 판단되면 CPU 또는 GPU 트랜스코더 중 보다 적합한 타입의 장치에 작업을 할당함으로써 연산 자원의 효율성을 높인다. 성능 평가 결과 듀얼 트랜스코딩 기법은 기존 방식 대비 최대 1.84배의 성능 향상을 이루었다. 결과적으로 단일 서버가 보다 많은 사용자의 요청을 QoS 유지 하에 제공할 수 있게 되었다.

• 한국통신학회논문지
• /
• 제28권12C호
• /
• pp.1209-1216
• /
• 2003

다양한 유무선 네트워크 환경상에서 실시간으로 보다 안정적이고 신뢰성 있는 주문형 비디오 및 실시간 멀티미디어 서비스론 사장자가 제공받기 위해서는 비디오 비트 스트림의 비트율을 다양한 채널의 대역폭에 맞추는 트랜스 부호화(transcoding) 과정이 필요하다. 본 논문에서는 트랜스 부호화 과정에서 움직임 벡터 추정시 매크로블록의 활동도와 방향성을 이용하여 계산랑을 줄이면서 화질을 보장하는 방법을 제안한다. 매크로블록의 활동도와 방향성의 임계치는 복호화 과정에서 추출한 중간정보 중에서 DCT 계수를 사용하여 비디오마다 적합하게 적용이 가능하다. 그리고 제안된 비디오 트랜스 부호화 알고리즘을 이용하여 H.263 환경에서 영상의 비트율 변환을 통한 실험으로 적절한 수준의 하질 유지와 계산량의 감소를 확인 할 수 있었다.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2002년도 ITC-CSCC -2
• /
• pp.1023-1026
• /
• 2002

In this paper, we propose an efficient transcoding of MPEG video. Transcoding is the process of converting a compressed video format to another different compressed video format. We propose an simple and efficient transcoding by requantization in which MPEG coded video at high bit-rate is converted into MPEG bitstream at lower bit-rate. To reduce a image quality degradation, we use HVS(Human Visual System) that is the effect that visibility of noise is less in high activity regions than in low activity regions. By using the effect, the part of image in high activity region is coarsely quantized without seriously degrading the image quality. Experimental results show that the proposed method can provide good performance.

• 대한전기학회논문지:시스템및제어부문D
• /
• 제51권10호
• /
• pp.460-467
• /
• 2002

This paper proposes new motion vector re-estimation techniques for transcoding of heterogenous video coders. The first case is for transcoder of coding format. The coding algorithm with B-picture like MPEG standards is transcoded into the different algorithm without B-picture like H.261 standards. The second case is for transcoder of spatial resolution reduction. It generates a downscaled video bitstream at the video source. The third case is the integrated method that previous two cases are needed simultaneously. These three transcoding algorithms are compared and verified though reconstructed image quality in PSNR, motion vector errors between optimal motion vectors of cascaded trans- coder and motion vectors estimated by proposed methods and the amount of operation.

• 한국통신학회논문지
• /
• 제33권7C호
• /
• pp.548-553
• /
• 2008

네트워크에 의해서 필요한 비트율로 줄이거나 터미널의 제한된 한계를 만족하기 위해서, 비디오 비트 스트림의 시간적 해상도를 줄이는 것이 사용된다. 본 논문에서는 압축된 비디오 스트림의 감소된 해상도 트랜스코딩의 문제점에 대해서 논의하고, 시간적 해상도 변환을 위한 트랜스코딩 기법을 논의한다. 트랜스코딩의 속도를 증가시키기 위해서, 비디오 트랜스코더는 보통 입력된 비디오 스트림으로부터의 움직임 벡터를 재사용한다. 본 논문에서는 코딩된 프레임의 높은 화질을 유지하기 위해서 강화된 움직인 재추정 기법을 제안한다. 실험결과는 감소된 프레임율을 가진 비디오 트랜스코더를 위한 저 복잡도를 가지면서도 성능이 개선된 것을 보여준다.

• Journal of information and communication convergence engineering
• /
• 제8권3호
• /
• pp.245-250
• /
• 2010

To support various bandwidth requirements for many kinds of devices such as PC, notebook, PDA, cellular phone, a transcoding proxy is usually necessary to provide not only adapting multimedia streams to the client by transcoding, but also caching them for later use. Due to huge size of streaming media, we proposed the 3 kinds of segmentation - PT-2, uniform, bitrate-based segmentation. And to reduce the CPU cost of transcoding video, we proposed the DQS service policy. In this paper, we simulate the combined our previous two researches that are bitrate-based segmentation and DQS(Degraded Quality Service) policy. Experimental results show that the combined policy outperforms companion schemes in terms of the byte-hit ratios and delay saving ratios.

• ETRI Journal
• /
• 제31권3호
• /
• pp.333-335
• /
• 2009

To match the requirements of heterogeneous mobile devices, video objects may be transcoded, which requires considerable CPU resources. Alternatively, multiple versions of the same video may be stored on servers, but this requires a lot of disk space. We formulate the trade-off between the versions that are stored on disk and the need for transcoding. We propose an optimal solution to this formulation based on dynamic programming. Experiments show that our scheme allows up to 68% more clients to be admitted than conventional schemes when a reasonable amount of storage is available.