《计算机应用研究》|Application Research of Computers

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

免费全文下载 (已被下载 次)  
获取PDF全文
作者 陈信新,王福林,宋莹莹
机构 东北农业大学 工程学院,哈尔滨 150030
统计 摘要被查看 次,已被下载
文章编号 1001-3695(2019)05-066-1592-05
DOI 10.19734/j.issn.1001-3695.2017.12.0837
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。
关键词 秧苗徒长; 状态监测; 视觉技术
基金项目 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)
本文URL http://www.arocmag.com/article/01-2019-05-066.html
英文标题 Rice seedling status recognition based on machine vision algorithm
作者英文名 Chen Xinxin, Wang Fulin, Song Yingying
机构英文名 College of Engineering,Northeast Agricultural University,Harbin 150030,China
英文摘要 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.
英文关键词 leggy seedlings; state monitoring; visual technology
参考文献 查看稿件参考文献
 
收稿日期 2017/12/26
修回日期 2018/2/6
页码 1592-1596
中图分类号 TP391.41
文献标志码 A