In this blog, I am sharing the final year project during my bachelor's which was Video Clone Detector using Hadoop which I and my partner successfully implemented under the guidance of our project guide from the university. So lets begin:-
Nowadays
internet is commonly used by the people as huge number of data is available on
the internet. This data is in the form of text, audio, video, etc. Out of this,
video is the one which is most trending among people because of its clear
expression as video depicts the concept more clearly compared to that of audio,
text which consumes a lot of time. But the problem is that the huge number of
videos already exists on the internet and also more and more videos are
generated every day. Due to which more storage space is required to store this
videos. Among these large number of videos, most of the videos are duplicated
copies of original ones. When the users browse for the video of his/her
interest then he/she has to go through this number of near duplicated copies.
Because of all these users time is wasted. To avoid this scenario, the need of
video copy detection is emerging rapidly.
Video
copy detection is to detect the copied video from the original ones by
comparing the content within them. The contents can be calculated by using
their hash values. It is difficult to achieve video copy detection using a
single machine approach. This problem can be solved using Hadoop which is a
distributed platform. In this MapReduce programming model is used, which uses
HDFS (Hadoop Distributed File System) as it is a distributed file storage
system. It is also use to achieve more reliable file storage. In order to
calculate and match the hash values of these large numbers of videos Hadoop is
used. Frame
Fusion based on Path Merging is the algorithm which is implemented using hadoop to detect the copied videos.
Step 1 :- Configure Hadoop
Step 2: Path merging technique
2.1 Admin Login
2.1.1 Uploading Video
Step 1: Register
An admin can
register into the system
Step 2: Login
An Admin can log
in to the system by inserting the correct credentials
Step 3: Upload
Video in Hdfs
An admin can upload
a video to HDFS by clicking on the browse button.
Step 4: Select
Video File
Select the video
file which you want to upload to the HDFS
Step 4:
Extraction of Frames
Frames would be
extracted from the video
Step 5: Generation of Brightness Sequence Values
Brightness Sequence Values of each frame of the video will be generated
Step 6: Generation
of Hash Sequence Values
Hash Sequence
Values of each frame of the videos will be generated
Step 7:
Uploading the generated data of the video to HDFS
The generated
data i.e Brightness sequence and Hash sequence values of each frames of videos
would be uploaded to HDFS
2.1.2 Checking Video
Step 1 :
Selecting the Video file
Select the video
file which you want to check by clicking on the check button
Step 2:
Extraction of Frames
Frames are
extracted from the video which is to be checked
Step 3:
Generation of Brightness Sequence Values
Brightness
Sequence values of each frame of the video that is to be check is generated
Step 4:
Generation of Hash Sequence Values
Hash sequence
values of the video that is to be checked is generated
Step 5: Result
It shows that
the video is original and it is 0% similar from the other videos
