基于深度学习和传统视觉相融合的农作物生长监测智能化关键技术研究
结题报告
批准号:
61962043
项目类别:
地区科学基金项目
资助金额:
38.0 万元
负责人:
郑媛
依托单位:
学科分类:
新型计算及其应用基础
结题年份:
2023
批准年份:
2019
项目状态:
已结题
项目参与者:
郑媛
国基评审专家1V1指导 中标率高出同行96.8%
结合最新热点,提供专业选题建议
深度指导申报书撰写,确保创新可行
指导项目中标800+,快速提高中标率
客服二维码
微信扫码咨询
中文摘要
借助先进的人工智能技术,由农业机器人代替农民监测农作物生长,既节省人力,又提高监测准确性。农业机器人配有多种相机,用于采集农作物图像。如何根据图像进行农作物生长监测成为一个关键问题,也是难点问题。本项目力求解决该难点问题所涉及的关键技术,具体研究内容包括:(1)农作物监控相机标定,提出一套准确、实用的相机标定方案,用于从图像中恢复农作物的生长特征参数;(2)基于图像的农作物分割,提出一种基于深度学习的农作物分割方法,提高对光照变化、背景干扰(例如杂草)和遮挡的鲁棒性;(3)基于图像的农作物生长评价模型,提出一种基于深度学习和传统视觉相融合的生长评价模型,该模型能够根据图像给出农作物生长信息,包括所处生长阶段、缺水、缺氮、缺光照情况。项目申请人在相机标定和基于图像的目标分析方面具有一定的科研积累;本项目的研究有助于实现农作物生长监测智能化,将为智慧农业提供坚实的理论基础和有力的技术支撑。
英文摘要
By means of artificial intelligence technology, the benefits of using agriculture robot to monitor the crop growing, instead of human beings, reside in the labor saving and improvement of surveillance accuracy. Agricultural robot is generally equipped with various cameras that are used to capture the crop images. How to monitor the crop growing using crop images is a key problem. This project aims to solve this problem, which contains the following research points: (1) calibration for crop surveillance camera, we will propose an accurate and practical camera calibration method, which is used to recover the feature parameters of crop growing; (2) image-based crop segmentation, we will propose a crop segmentation model based on deep learning, which is able to improve the robustness against to illumination variation, background disturbance (i.e. weed) and occlusion; (3) image-based evaluation model for crop growing, we will propose a crop growing evaluation model based on the combination of deep learning and traditional vision methods, which is able to provide the crop growing information, including growth phase, moisture condition, nutrition condition and lighting condition. The project applicant has done some research on camera calibration and image-based object analysis. This project is helpful to realize intelligent crop growing surveillance and will provide a solid theoretical foundation and an effective technical support for intelligent agriculture.
期刊论文列表
专著列表
科研奖励列表
会议论文列表
专利列表
DOI:10.1109/tcsvt.2023.3293166
发表时间:2024-02
期刊:IEEE Transactions on Circuits and Systems for Video Technology
影响因子:8.4
作者:Lei Chen;Huhe Dai;Yuan Zheng
通讯作者:Lei Chen;Huhe Dai;Yuan Zheng
交通场景下基于视频的智能监控分析关键技术研究
  • 批准号:
    61502119
  • 项目类别:
    青年科学基金项目
  • 资助金额:
    21.0万元
  • 批准年份:
    2015
  • 负责人:
    郑媛
  • 依托单位:
国内基金
海外基金