...
  • 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Cold regions science and technology >Determination of the lower boundary of a rotating ice patch for ice thickness estimation using image convolution and machine learning
【24h】

Determination of the lower boundary of a rotating ice patch for ice thickness estimation using image convolution and machine learning

机译：使用图像卷积和机器学习确定用于估计冰厚的旋转冰块的下边界

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

As the number of trips along the Arctic route has increased, the safe navigation of icebreakers on this route has become vital. One of the key factors that affect the stability of an icebreaker is the ice thickness in the route. The ice thickness, if measured in real time during navigation, can aid in the construction of an ice map or can be used for the structural analysis of the icebreaker after the voyage. However, the measurement of ice thickness during the voyage is challenging. In this study, a method to improve the measurement of ice thickness from an image is introduced. To measure the ice thickness accurately and quickly, a method to identify the lower boundary curve of the cross-section of an ice patch is developed. Pixel-based constraints are determined by considering the characteristics of color variation in an image. In this method, the lower boundary curve is determined by using image convolution and a specific filter, and the performance efficiency is enhanced through a machine learning technique. The results suggest that an accumulating learning process can increase the recognition rate.
机译：随着沿北极路线旅行次数的增加，在该路线上安全破冰船的航行变得至关重要。影响破冰船稳定性的关键因素之一是航线上的冰层厚度。如果在航行期间实时测量冰的厚度，则可以帮助构造冰图，或者可以用于航行后破冰船的结构分析。然而，在航行期间冰厚度的测量是具有挑战性的。在这项研究中，介绍了一种改进从图像中测量冰厚的方法。为了准确而快速地测量冰的厚度，开发了一种识别冰块横截面的下边界曲线的方法。通过考虑图像中颜色变化的特征来确定基于像素的约束。在这种方法中，下边界曲线通过使用图像卷积和特定的滤波器确定，并且通过机器学习技术提高了性能效率。结果表明，不断积累的学习过程可以提高识别率。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号