GPU-based mass-multimedia-data-oriented MapReduce platform

A multimedia data and platform technology, applied in the field of MapReduce platform, to achieve the effect of improving calculation speed, ensuring calculation accuracy, and high-performance processing

Inactive Publication Date: 2014-04-02
TONGJI UNIV
View PDF3 Cites 13 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Faced with the exponential growth of video information, especially the explosive development of online video, traditional computing and storage models are difficult to analyze and process these massive data information.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • GPU-based mass-multimedia-data-oriented MapReduce platform
  • GPU-based mass-multimedia-data-oriented MapReduce platform
  • GPU-based mass-multimedia-data-oriented MapReduce platform

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0016] The present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0017] Such as Figure 1-3 Shown, a kind of MapReduce platform based on GPU facing massive multimedia data utilizes computer clusters to realize the computation processing to image / video retrieval task, is provided with a plurality of CPUs and GPUs in each computer cluster, it is characterized in that, described The platform is built on CUDA and HDFS, including a platform driver 1 and a working sub-module 2. The platform driver 1 is responsible for managing the software and hardware resources of the platform, and controlling the workflow of the platform. Its main work includes task startup and segmentation. , scheduling, fault-tolerant processing, etc.; the working sub-module 2 is the realization of the main image and video retrieval processing algorithms such as feature point extraction, clustering, etc., and undertakes the heaviest computin...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention relates to a GPU-based mass-multimedia-data-oriented MapReduce platform comprising a platform driver and an operating sub-module. The platform driver divides an image / video retrieval task into a plurality of Map tasks by a MapReduce calculation model. Data of Map tasks is stored in an HDFS (hadoop distributed file system). When each Map task starts, task data are acquired by a file list transmitted in by the platform driver, a specific calculation task is assigned to the operating sub-module for calculation, a task scheduler in the operating sub-module assigns the task to the GPU or CPU for processing, calculation-required data is acquired through a local library libhdfs.so of the HDFS during the calculation process, and the data calculated is directly written into the HDFS. Compared to the prior art, the GPU-based mass-multimedia-data-oriented MapReduce platform has the advantages that mass multimedia data can be processed under high performance, calculation speed is greatly increased and calculation precision can also be guaranteed.

Description

technical field [0001] The invention relates to the technical fields of massive data processing and high-performance computing processing, in particular to a GPU-based MapReduce platform for massive multimedia data. Background technique [0002] When the information age entered Web2.0, with the rise of multimedia original interactive system, the popularity of new media such as network multimedia and mobile multimedia, and the popularity and popularization of portable smart terminal devices (such as: IPhone, IPad, notebook, etc.), The amount of multimedia (such as video, images, etc.) on the Internet is showing a massive and explosive growth. Massive pictures and videos are transmitted on the Internet, and searching and viewing rich picture and video resources through the Internet has become an important way for the majority of netizens to obtain information. In the face of massive multimedia data, how to effectively organize, manage and search it has become an urgent task, ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): G06F17/30G06F9/50
CPCG06F9/5027
Inventor 王瀚漓肖波王雷朱冯贶天
Owner TONGJI UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap