...
首页> 外文期刊>Agrociencia >Identification of eroded zones with digital images using artificial neural networks.
【24h】

Identification of eroded zones with digital images using artificial neural networks.

机译:使用人工神经网络通过数字图像识别侵蚀区域。

获取原文
获取原文并翻译 | 示例
           

摘要

Hydric erosion is a phenomenon that has an impact on crop production; but its evaluation is very difficult due to its constant dynamics, cost and time required. In this study, a methodology is proposed to identify hydric erosion using a digital image analysis. The study was conducted Santa Catarina del Monte, Texcoco, Mexico. The proposed methodology consists of using a backpropagation artificial neural network as pixel classifier using the Levenberg-Marquardt training algorithm. The network is capable of identifying white tepetate with an error of 2.5%; tepetate in transition from white to yellow with 16%; trees with 13.5%, and soils covered with arvenses with 7.1%. For yellow tepetates, the error was up to 3560%; the same trend was observed in the identification of gullies.
机译:水土流失是一种影响农作物产量的现象。但是由于其不断变化的动态,所需的成本和时间,因此对其进行评估非常困难。在这项研究中,提出了一种使用数字图像分析来识别水蚀的方法。该研究是在墨西哥Texcoco的Santa Catarina del Monte进行的。所提出的方法包括使用反向传播人工神经网络作为像素分类器,并使用Levenberg-Marquardt训练算法。该网络能够识别2.5%的白色tepetate; tepetate从白色到黄色的过渡,占16%;树木占13.5%,土壤覆盖着植物覆盖物占7.1%。对于黄色的tepetates,其误差高达3560%。在识别沟渠中观察到了相同的趋势。

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