Scholarlink

In this paper, a novel remote sensing image object detection algorithm is proposed by combining constraint energy minimization (CEM) and noise tolerance zeroing neural network (NTZNN). This algorithm combines traditional image processing methods with recursive neural network (RNN) and proposes a multi scene usable NTZNN-CEM object detection model for hyperspectral and RGB remote sensing images. Finally, through numerical simulation experiments and remote sensing image object detection experiments, it has been proven that the NTZNN-CEM algorithm has advantages such as fast detection speed and strong robustness, providing a new visual approach for improving the advanced perception of robots in complex environments.

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

Recursive Neural Network: Small Target Detection in Remote Sensing Images

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Changlin Yu,
  • Juchao Zhang,
  • Zhongyu Sun,
  • Zaixiang Pang,
  • Changxian Xu,
  • Zhongbo Sun

摘要

In this paper, a novel remote sensing image object detection algorithm is proposed by combining constraint energy minimization (CEM) and noise tolerance zeroing neural network (NTZNN). This algorithm combines traditional image processing methods with recursive neural network (RNN) and proposes a multi scene usable NTZNN-CEM object detection model for hyperspectral and RGB remote sensing images. Finally, through numerical simulation experiments and remote sensing image object detection experiments, it has been proven that the NTZNN-CEM algorithm has advantages such as fast detection speed and strong robustness, providing a new visual approach for improving the advanced perception of robots in complex environments.