基于机器视觉算法的水稻秧苗状态识别  

Rice seedling status recognition based on machine vision algorithm

作　　者：陈信新 王福林[1] 宋莹莹 Chen Xinxin;Wang Fulin;Song Yingying(College of Engineering, Northeast Agricultural University, Harbin 150030, China)

机构地区：[1]东北农业大学工程学院,哈尔滨150030

出　　处：《计算机应用研究》2019年第5期1592-1596,共5页Application Research of Computers

基　　金：国家自然科学基金资助项目(31071331);公益性行业(农业)专项课题资助项目(201503116-04)

摘　　要：根据秧苗发育形态调节育秧环境中的光照、温度、湿度能够避免徒长,这对水稻优质高产十分重要。针对水稻秧苗在恒温箱封闭式育秧环境下因监测不及时导致的徒长情况,提出了一种基于机器视觉算法识别秧苗发育状态的方法。采用恒温箱对东农426和东农428水稻进行育秧研究,利用协方差聚类算法对秧苗RGB彩色图像进行分割,再选用连续腐蚀开操作结合Hough变换进行预处理。根据提取与秧苗徒长直接相关的形态指标参数信息,如株高、叶面积、着生角、生长速率,进行曲线拟合,最后将结果显示在软件界面上。实验结果表明,该方法可以正确地识别秧苗并准确提取形态指标参数,准确率为87.5%,秧苗形态参数识别误差不超过7%,适用于封闭式育秧环境中对秧苗的监测,该方法为研究水稻工厂化立体育秧提供了有效参考。It is very important to improve the quality and high yield of rice by adjusting the light,temperature and humidity in the environment of seedling development.This paper proposed a method for recognizing seedling development status based on machine vision algorithm for rice seedlings under the condition that the monitoring of rice seedlings was out of time caused by monitoring.It studied on the seedling of Dongnong426 and Dongnong428 by rice thermotank,on seedling RGB color image segmentation using the covariance clustering algorithm,then used continuous corrosion pretreatment open operation combined with Hough transform.According to the information extraction and morphological parameters of leggy seedlings directly related,such as plant height,leaf area,horned,growth rate,curve fitting,the results would be displayed on the interface of the software.The test results show that the method can correctly identify and accurately extract the morphological parameters of seedlings,the accuracy rate is 87.5%,the morphological parameter identification error is less than 7%,this method provides an effective reference for the research of rice seedling sports factory.

关 键 词：秧苗徒长 状态监测 视觉技术 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]