Scholarlink

<p>Researchers have been actively investigating the statistics about Land Use/Land Cover worldwide for several decades. This research examines the Javadi Hills in India, a region known for its natural significance, rich forest resources, and essential water content within the forest and non-forest landscapes. The LISS-III satellite imageries were used in this work to explore the Land Use/Land Cover statistics in water bodies, vegetation, and bare soil for the periods of past (2012, 2015, 2018, 2021), present (2024), and future (2027), with the selected timeframe ensuring a consistent three-year gap between each period to facilitate data availability and forecast future trends. We implemented the novel Fuzzy Shuffled Frog Leaping Optimization-based Convolutional Long Short-Term Memory (FSFLO-ConvLSTM) model, which integrates the Normalized Difference Water Index (NDWI) as a fuzzy membership function and spatial mapping variable. This method significantly enhances the classification accuracy (96.84%) and prediction accuracy (95.73%). The importance of this research was using NDWI to improve accuracy and create a spatial map, which helps in understanding changes in vegetation and water bodies. The overall statistics of past (2012–2021), present (2024), and future (2027) Land Use/ Land Cover assist the urban designers, government representatives, and concerned forestry officers to safeguard the environment, especially the vegetation and water bodies.&#xa0;</p>

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

Fuzzy Shuffled Frog Leaping Optimization-based enhanced ConvLSTM for Land Use/ Land Cover Prediction

  • 期刊论文

  • 发表日期：

    2025年1月27日

查看全文
  • Sam Navin MohanRajan,
  • Agilandeeswari Loganthan,
  • Prabukumar Manoharan

摘要

Researchers have been actively investigating the statistics about Land Use/Land Cover worldwide for several decades. This research examines the Javadi Hills in India, a region known for its natural significance, rich forest resources, and essential water content within the forest and non-forest landscapes. The LISS-III satellite imageries were used in this work to explore the Land Use/Land Cover statistics in water bodies, vegetation, and bare soil for the periods of past (2012, 2015, 2018, 2021), present (2024), and future (2027), with the selected timeframe ensuring a consistent three-year gap between each period to facilitate data availability and forecast future trends. We implemented the novel Fuzzy Shuffled Frog Leaping Optimization-based Convolutional Long Short-Term Memory (FSFLO-ConvLSTM) model, which integrates the Normalized Difference Water Index (NDWI) as a fuzzy membership function and spatial mapping variable. This method significantly enhances the classification accuracy (96.84%) and prediction accuracy (95.73%). The importance of this research was using NDWI to improve accuracy and create a spatial map, which helps in understanding changes in vegetation and water bodies. The overall statistics of past (2012–2021), present (2024), and future (2027) Land Use/ Land Cover assist the urban designers, government representatives, and concerned forestry officers to safeguard the environment, especially the vegetation and water bodies.