Cloud based Multimedai Application for Handheld devices: Using Python wrapper

Abstract

the exponential growth and availability of the multimedia information had become more popular in
research and technology field. The huge development of the multimedia information and obtainable bandwidth there
are more demand for video retrieval system in many domains like education, entertainment, healthcare, and social
media etc. The amount of data becoming very heavy, where the retrieval of the content become an issue. To overcome
this issue the indexing is used. The indexing collects the data, parses the data and stores it to simplify for fast and
correct information retrieval. By using cloud based multimedia computing technology (CMM) we can store and
compute or process the multimedia application data on cloud in distributed manner. To extract the key frames from
the video file, the open source tool FFMPEG and MPEG-7 descriptor is used with the python 3.2 wrapper and java
programming language, MFCC is used to extract the audio feature, Apache Solr is used as a search engine where the
index files are updated using CURL Protocol. HTML5 is used for a graphical user interface where the users can
request the query, and the video data will be transferred in JSON format over the network, converting the video will be
also done within the cloud, so that computational resource required in cloud not on handheld devices. As video data
will be accessed from the private cloud since there will be no need of storage space for video data on handheld device.
The proposed system will give 97% for trained and 94% for non-trained videos and shows the 193ms to 716ms
retrieval time.