首页> 外文会议>Iberian Conference on Information Systems and Technologies >Inteligent system to identify oil palm crop units from multispectral aerial images: Identification of multispectral patterns
【24h】

Inteligent system to identify oil palm crop units from multispectral aerial images: Identification of multispectral patterns

机译:从多光谱航拍图像识别油棕作物单位的智能系统:多光谱模式的识别

获取原文

摘要

Globally, the oil palm industry leads the supply of oils and fats, and is a very dynamic sector thanks to its wide use and applications in different products. However, some studies have shown that oil palm cultivation causes enormous damage to the environment by the destruction of existing forests, or by excessive use of fertilizers and pesticides, making crude palm oil uncompetitive and strong constraints for its implementation worldwide. To improve the production of crude palm oil, assisted pollination plays a fundamental role in reducing the use of fertilizers. This is why this article presents a flexible Fuzzy ARTMAP model, for the segmentation and identification of oil palm crop units from the characterization of multispectral aerial images. The results show that the proposed model was able to identify complete seed units from polar patterns obtained from the segmentation using grayscale images, as well as the identification of sowing unit that are susceptible to be pollinated by the characterization of each unit of planting using three vegetation index such as: NDVI (Normalized Difference Vegetation Index), GNDVI (Green NDVI) y RVI (Red Vegetation Index).
机译:在全球范围内,油棕行业引领着油脂的供应,并且由于其在不同产品中的广泛使用和应用,它是一个非常活跃的行业。但是,一些研究表明,油棕种植会因破坏现有森林或过度使用化肥和杀虫剂而对环境造成巨大破坏,从而使粗棕榈油失去竞争力,并成为在全球范围内实施的强大限制。为了提高粗棕榈油的产量,辅助授粉在减少肥料使用方面起着根本性的作用。这就是为什么本文提出了一种灵活的Fuzzy ARTMAP模型,用于通过多光谱航拍图像的特征来分割和识别油棕作物单元。结果表明,所提出的模型能够使用灰度图像从分割得到的极地模式中识别出完整的种子单元,并且能够识别出通过使用三种植被对每个种植单元进行表征而易于授粉的播种单元。指数,例如:NDVI(归一化植被指数),GNDVI(绿色NDVI)和RVI(红色植被指数)。

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