Detection of vegetation in environmental repeat photography: A new algorithmic approach in data science

Asim Khan, Anwaar Ul-Haq, Mobeen ur Rehman, Randall W Robinson

Research output: Book chapter/Published conference paperConference paper

Abstract

Environment change being one of the major issues in today’s world
needs special attention of the researchers. With the advancement in computer
vision researchers are equipped enough to come up with algorithms accomplishing
automated system for environment monitoring. This paper proposes an algorithm
which can be used to observe the change in vegetation utilizing the images of a
particular site. This would help the environment experts to put on their efforts in a
right direction and right place to improve the environment situation. The proposed
algorithm registers the image so that comparison can be carried out in an accurate
manner using single framework for all the images. Registration algorithm aligns the
new images with the existing images available in the record of the same particular
site by performing transformation. Registration process is followed by segmentation
process which segments out the vegetation region from the image. A novel approach
towards segmentation is proposed which works on the machine learning based
algorithm. The algorithm performs classification between vegetation patches and
non-vegetation patches which equips us to perform segmentation. The proposed
algorithm showed promising results with F-measure of 85.36%. The segmentation
result leads us to easy going calculation of vegetation index. Which can be used to
make a vegetation record regarding particular site.
Original languageEnglish
Title of host publicationStatistics for Data Science and Policy Analysis
EditorsAzizur Rahman
PublisherSpringer Singapore
Chapter11
Pages145-157
Number of pages13
Edition1
ISBN (Electronic)9789811517358
ISBN (Print)9789811517341
Publication statusPublished - 2020
EventThe 2nd Applied Statistics and Policy Analysis Conference: ASPAC2019 - Charles Sturt University, Wagga Wagga, Australia
Duration: 05 Sep 201906 Sep 2019
http://csusap.csu.edu.au/~azrahman/ASPAC2019/
http://csusap.csu.edu.au/~azrahman/ASPAC2019/Program%20draft.pdf?attredirects=0&d=1 (program)
http://csusap.csu.edu.au/~azrahman/ASPAC2019/ASPAC2019_Refereed_Book%20of%20Abstracts.pdf?attredirects=0&d=1 (book of abstracts)
https://ebookcentral.proquest.com/lib/CSUAU/detail.action?docID=6152166 (proceedings)

Conference

ConferenceThe 2nd Applied Statistics and Policy Analysis Conference
Abbreviated titleEffective policy through the use of big data, accurate estimates and modern computing tools and statistical modelling
CountryAustralia
CityWagga Wagga
Period05/09/1906/09/19
OtherProceedings due for publication May 2020 https://www.springer.com/gp/book/9789811517341
Internet address

Fingerprint Dive into the research topics of 'Detection of vegetation in environmental repeat photography: A new algorithmic approach in data science'. Together they form a unique fingerprint.

  • Cite this

    Khan, A., Ul-Haq, A., ur Rehman, M., & Robinson, R. W. (2020). Detection of vegetation in environmental repeat photography: A new algorithmic approach in data science. In A. Rahman (Ed.), Statistics for Data Science and Policy Analysis (1 ed., pp. 145-157). Springer Singapore.