Scholarlink

Modified Normalized Difference Water Index (MNDWI) is a spectral index used to analyze and map the aquaculture ponds from the Sentinel-2 satellite images that are available in Google Earth Engine (GEE). Monitoring and census of aquacultural activity is conventionally undertaken by field surveys that are time-consuming and expensive. The large-scale identification of aquaculture ponds has also been greatly enhanced by cloud computing platforms like GEE. The main objective of the study is that the machine learning classifier Random Forest (RF) was employed in GEE-based mapping investigations of aquaculture ponds. The study has shown that the Random Forest classifier is capable of extracting aquaculture pond spatial distribution with excellent overall accuracy, exceeding 90%. The locations derived from Google Earth Imagery are found to be consistent with the corresponding aquaculture ponds depicted on the resultant map obtained from the RF classifier.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Extracting Aquaculture Ponds from Natural Water Surfaces Around Inland Lakes Through GEE Datasets: A Case-Study of River Krishna Delta in South India

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Saravanan Subbarayan,
  • Ramanarayan Sankriti,
  • Dakshinamurthy Ravi Nemani,
  • Divya Ranganathachar,
  • E. Arivoli,
  • Niraimathi Janardhanam

摘要

Modified Normalized Difference Water Index (MNDWI) is a spectral index used to analyze and map the aquaculture ponds from the Sentinel-2 satellite images that are available in Google Earth Engine (GEE). Monitoring and census of aquacultural activity is conventionally undertaken by field surveys that are time-consuming and expensive. The large-scale identification of aquaculture ponds has also been greatly enhanced by cloud computing platforms like GEE. The main objective of the study is that the machine learning classifier Random Forest (RF) was employed in GEE-based mapping investigations of aquaculture ponds. The study has shown that the Random Forest classifier is capable of extracting aquaculture pond spatial distribution with excellent overall accuracy, exceeding 90%. The locations derived from Google Earth Imagery are found to be consistent with the corresponding aquaculture ponds depicted on the resultant map obtained from the RF classifier.