测绘通报 ›› 2020, Vol. 0 ›› Issue (7): 76-81.doi: 10.13474/j.cnki.11-2246.2020.0218

• 学术研究 • 上一篇    下一篇

利用无人机遥感监测农作物种植面积

任泽茜3, 丁丽霞1,2,3, 刘丽娟1,2,3, 谢锦莹2,3, 敖伊颍3, 张继艳3, 何嘉莹3   

  1. 1. 浙江农林大学省部共建亚热带森林培育国家重点实验室, 浙江 杭州 311300;
    2. 浙江农林大学 浙江省森林生态系统碳循环与固碳减排重点实验室, 浙江 杭州 311300;
    3. 浙江农林大学环境与 资源学院, 浙江 杭州 311300
  • 收稿日期:2019-10-11 出版日期:2020-07-25 发布日期:2020-08-01
  • 通讯作者: 丁丽霞。E-mail:dlxlxy@126.com E-mail:dlxlxy@126.com
  • 作者简介:任泽茜(1997-),女,研究方向为无人机遥感在农业监测中的应用。E-mail:3475968541@qq.com
  • 基金资助:
    国家自然科学基金(31870619)

Crop acreage monitoring based on UAV image

REN Zexi3, DING Lixia1,2,3, LIU Lijuan1,2,3, XIE Jinying2,3, AO Yiying3, ZHANG Jiyan3, HE Jiaying3   

  1. 1. State Key Laboratory of Subtropical Silviculture, Zhejiang A&F University, Hangzhou 311300, China;
    2. Key Laboratory of Carbon Cycling Forest Ecosystems and Carbon Sequestration of Zhejiang Province, Zhejiang A&F University, Hangzhou 311300, China;
    3. School of Environmental and Resource Sciences, Zhejiang A&F University, Hangzhou 311300, China
  • Received:2019-10-11 Online:2020-07-25 Published:2020-08-01

摘要: 南方平原耕地具有地块破碎、农作物种植品种多且空间分布混杂程度高等特点,运用传统的遥感技术方法精确监测农作物面积较为困难。无人机航拍具有拍摄时间灵活、空间分辨率高、成本低等优势,为解决这一难题提供了有利途径。本文通过地面样地调查,获取杭州市余杭区瓶窑镇农作物样地的位置及种植品种数据,利用面向对象的多尺度分割方法与随机森林的分类方法对无人机航拍数据进行分割、分类,深入挖掘高分辨率遥感数据信息,用于提取农作物种植品种及其空间分布信息,实现高精度的农作物种植面积遥感监测,推进无人机遥感在农业中的深入应用,提高农业遥感应用效益。

关键词: 无人机, 高分辨率图像, 多尺度分割, 随机森林分类, 农作物面积

Abstract: The cultivated land in the southern plain is fragmented, varieties of crops planted and the spatial distribution is mixed, so it is difficult to accurately monitor the area of crops by using traditional remote sensing technology. Drones aerial photography has advantages providing a beneficial way to solve this problem which is flexible for shooting time, high spatial resolution and low cost, et al. In this paper, drones aerial photography data is used to monitor the ground sample plots in Pingyao town, Yuhang district, Hangzhou city, to obtain the the location and variety data of the crops. The object-oriented multi-scale segmentation method and the random forest classification method are used to segment and classify the aerial data of the drone, and the high-resolution remote sensing data information is used to extract the crop varieties and their spatial information to achieve high-precision crops area, to promotthe deeply application of UAV remote sensing in agriculture.

Key words: UAV, high-resolution image, multi-scale segmentation, random forest classificaton, crop area

中图分类号: