Scholarlink

Automated terrain recognition is a vital and essential task. It has a great application in the field of agricuture, weather, disaster etc. This paper used CNNs and other transfer learning models to categorize terrains into four distinct classes: grassy, marshy, sandy, and rocky. Better performance was given by ResNet50 with an accuracy of 99.2%. Other models also exhibit similar performance but lesser accuracy. This research could help modelling real-time terrain recognition system for various objectives.

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

Machine Learning Based Terrain Recognition for Agricultural and Environmental Development

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Siva Sai Susmitha Katta,
  • Greeshmitha Vavilapalli,
  • Junali Jasmine Jena,
  • Suresh Chandra Satapathy,
  • Mahendra Kumar Gourisaria,
  • Amiya Ranjan Panda

摘要

Automated terrain recognition is a vital and essential task. It has a great application in the field of agricuture, weather, disaster etc. This paper used CNNs and other transfer learning models to categorize terrains into four distinct classes: grassy, marshy, sandy, and rocky. Better performance was given by ResNet50 with an accuracy of 99.2%. Other models also exhibit similar performance but lesser accuracy. This research could help modelling real-time terrain recognition system for various objectives.