利用先验知识的小麦生长环境下杂草识别研究  

Weed Recognition in Wheat Growing Environment Using Prior Knowledge

作　　者：胡晓斌 HU Xiao-bin(School of Environmental and Geomatics Engineering,Suzhou University,Suzhou,Anhui 234000;College of Surveyingand Mapping,Wuhan University,Wuhan,Hubei 430079)

机构地区：[1]宿州学院环境与测绘工程学院,安徽宿州234000 [2]武汉大学测绘学院,湖北武汉430079

出　　处：《安徽农业科学》2018年第25期176-178,共3页Journal of Anhui Agricultural Sciences

基　　金：国家自然科学基金项目(41271456);安徽省人文社科重点项目(SK2017A0728);宿州市社科规划课题(SZ2017A106);宿州学院博士基金项目(2014jb03)

摘　　要：随着土地流转机制的成熟和完善,以及由传感器、通信和计算机等技术组成的物联网系统的发展,为农业现代化管理提供了制度和技术的保障,其中农作物生长环境下杂草的识别和清除是家庭农场管理的重要组成部分。首先利用家庭农场作物先验几何特性,拟合小麦正常生长条件下的中心线,根据小麦种植时固定的行间距,识别小麦生长环境下的杂草,对比实地调查的统计数据,该研究算法识别杂草的正确率达到92.8%,可为家庭农场管理装备提供技术支撑。With the maturity and perfection of the land transfer mechanism , as well as the development of the Internet of things, such as sen-sor , communication and computer technology, the system and technology are provided for the management of agricultural modernization. First, we use the prior geometric characteristics of family farm crops to fit the central line of wheat under normal growth conditions ,identify weeds under the fixed row spacing of wheat, and compare the statistical data of field survey. The accuracy of this algorithm is 92. 8% , which can be provided for family farm management equipment.

关 键 词：土地流转 先验知识 小麦 杂草识别 

分 类 号：S24[农业科学—农业电气化与自动化] TP39[农业科学—农业工程]