图像处理技术在田间杂草识别中应用研究  

Application of image processing technology in weed recognition in field

作　　者：任全会 杨保海 Ren Quanhui;Yang Baohai(Henan Province Smart Education and Intelligent Technology Application Engineering Technology Research Center,Zhengzhou,451460,China;College of Physics and Electronic Engineering,Guangxi Normal University for Nationalities,Chongzuo,532200,China)

机构地区：[1]河南省智慧教育与智能技术应用工程技术研究中心,郑州市451460 [2]广西民族师范学院物理与电子工程学院,广西崇左532200

出　　处：《中国农机化学报》2020年第6期154-158,共5页Journal of Chinese Agricultural Mechanization

基　　金：国家自然科学基金项目(41001251);河南省科技厅科技攻关项目(182102918085);广西教育厅科技项目(2019KY0784);广西民族师范学院科学研究项目(2019FG008)。

摘　　要：田间杂草识是监测田间生物环境系统状态的重要方法,使用传统方法在进行田间杂草识别时普遍存在识别率低的问题。鉴于此,文章提出了使用图像处理技术进行田间杂草的精确识别。首先对田间图像进行预处理,充分利用Canny算子的特点进行图像边缘检测,同时把四个重要的特征参数计算出来,然后通过遗传算法的特点构造出田间杂草识别模型。文章以玉米幼苗、刺儿菜、荠菜、婆婆纳、香附子、田旋花、小飞蓬、粘毛卷耳等八种植物作为样本,使用本文方法在相同样本数量和条件下与神经网络算法进行了测试和对比分析,八种植物的识别平均错误率分别为2.8%、2.6%、2.9%、3.0%、2.4%、3.1%、2.7%、2.5%,通过对测试结果进行分析可以看出,和传统的方法相比,平均识别错误率大大降低,具有很强的实用价值。Field weed identification is an important method to monitor the status of the field biological environment system.The traditional method has the problem of low recognition rate.In view of this,this paper proposes the use of image processing technology to accurately identify weeds in the field.Firstly,the field image is preprocessed,and the edge of the image is detected by Canny operator.At the same time,four important characteristic parameters are calculated,and then the field weed recognition model is constructed by the characteristics of genetic algorithm.In this paper,eight kinds of plants,such as corn seedling,bayonet,shepherd’s purse,mother-in-law,xiangfuzi,tianxuanhua,xiaofeipeng,curly ears and so on,are used as samples.The average recognition error rate of eight kinds of plants is 2.8%,2.6%,2.9%,3.0%,2.4%,3.1%,2.7%and 2.5%respectively,through the analysis of the test results,we can see that compared with the traditional method,the average recognition error rate is greatly reduced,which has a strong practical value.

关 键 词：CANNY算子 杂草识别 图像处理 遗传算法 特征参数 边缘检测 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]