基于时间序列图像的玉米植株干旱胁迫表型检测方法

发布时间：2018-01-03 08:07

本文关键词：基于时间序列图像的玉米植株干旱胁迫表型检测方法　出处：《农业工程学报》2016年21期 　论文类型：期刊论文

【摘要】：为了快速准确获取植物表型数据,该文提出了一种基于时间序列多图像的玉米植株干旱胁迫表型检测方法。首先,以固定时间间隔获取干旱胁迫条件下玉米植株一天内的图像数据,采用统计学习方法对图像序列进行初始背景分割,非局部均值(non-local mean,NLM)滤波对初始背景分割进一步求精,去除变化光照条件对图像序列分割的影响。在植株器官分割阶段,融合植株形态先验知识,通过逐行扫描和邻域数量判断拆分植株器官像素。最后,提出了叶片角度比值的干旱胁迫表型参数计算方法。结果显示,耐旱植株的叶片角度比值一天内从0.63上升到0.70后再降到0.65,而非耐旱植株的叶片角度比值维持在0.65左右,随时间变化不大。研究结果为阐明玉米植株耐旱表型参数与环境参数的协同变化关系,以及不同基因型玉米植株环境应激反应的细微形态变化提供方法和技术解决方案。
[Abstract]:In order to quickly and accurately obtain the plant phenotypic data, this paper proposes a corn plant drought stress time series multi image detection method based on phenotypic. Firstly, the image data acquisition of maize plants under drought stress in a day at a fixed time interval, using the statistical learning method of initial background image sequence segmentation, non local means (NON-LOCAL mean, NLM) to filter the initial background segmentation further refinement, removing the influence of illumination changes on the image sequence segmentation. Segmentation stage in plant organs, fusion plant shape prior knowledge, through the progressive scan and determine the number of split neighborhood plant organ pixels. Finally, puts forward the calculation method of the blade angle ratio of phenotypic parameters under drought stress. The results show that the blade angle of the ratio of drought tolerant plants within a day rose from 0.63 to 0.70 after dropping to 0.65, rather than the blade angle of drought tolerant plants The ratio is maintained at about 0.65, which has little change with time. The results provide a way to elucidate the synergistic relationship between drought tolerance phenotypic parameters and environmental parameters of maize, and to provide methods and technical solutions for different morphological changes of plant stress responses.

【作者单位】：北京农业信息技术研究中心;数字植物北京市重点实验室;
【基金】：国家863计划课题(2013AA102404) 农业部行业科技计划项目(201203026) 国家自然科学基金项目(31501226) 北京市农林科学院数字植物科技创新团队(JNKYT201604)
【分类号】：S513;TP391.41
【正文快照】： 王传宇,郭新宇,杜建军,温维亮,吴升.基于时间序列图像的玉米植株干旱胁迫表型检测方法[J].农业工程学报,2016,32(21):189-195.doi:10.11975/j.issn.1002-6819.2016.21.025 http://www.tcsae.orgWang Chuanyu,Guo Xinyu,Du Jianjun,Wen Weiliang,Wu Sheng.Maize plant drought

【相似文献】