应用数字图像分析技术进行棉花氮素营养诊断的研究  

Diagnosis of cotton N status using digital image analysis technique

作　　者：王娟[1] 雷咏雯[1] 张永帅[1] 郭金强[2] 危常州[1] 

机构地区：[1]石河子大学新疆兵团绿洲生态农业重点实验室,石河子832003 [2]石河子市气象局,石河子832000

出　　处：《中国生态农业学报》2008年第1期145-149,共5页Chinese Journal of Eco-Agriculture

基　　金：国家自然科学基金项目(30560069);农业部“948”重大专项(2003-Z53)资助

摘　　要：本文利用图像分析技术并结合常规观测手段,研究应用图像分析技术诊断棉花氮素营养状况的可行性及获取的光谱参数与表征棉花氮素营养状况的生物学参数之间的关系。结果表明:棉花在不同时期特征光谱参数与棉花含氮量及叶片含氮量呈显著相关,其中盛蕾期棉花全氮含量与光谱参数的相关性最好,在盛花期棉花叶片含氮量与光谱参数的相关系数最高,G/(G+R+B)可作为氮素营养诊断的指标。在棉花全生育期内,地面覆盖度与棉花叶面积指数、生物量及吸氮量呈显著相关,在出苗至盛花期之间达极显著相关。经检验,地面覆盖度可很好地预测棉花的叶面积指数、生物量及吸氮量,相对误差分别为26.2%、3.46%和3.37%。This paper is an investigation of the possibility of N nutrition diagnosis in cotton plant, using digital image analysis of digital camera shots, and researches into the relationships between spectral and biology parameters, which could denote cotton N status. The results show a significant linear correlation between feature spectral parameters and total N concentration in cotton plant. The same holds true for the relationship of cotton leaf N concentration. The best relationship between spectral parameters and total N concentration occurs during full-budding stage. However, at full-flowering stage, regression coefficient between spectral parameters and cotton leaf N concentration is the best. G/（ G ＋ R ＋ B） parameter can be considered an important indictor of N status diagnosis. Results of the experiment show that ground-canopy coverage percent and LAI,biomass and accumulation of absorbed N, significantly correlate for the entire cotton growth stage. From sprouting to full-flowering, the correlation hits a high level of significance（P 〈 0.01 ）. At last the test results indicate that ground-canopy coverage percent has a better estimation precision for LAI, biomass and accumulated absorbed N, and with a relative error percent of 26.2% ,3.46% and 3.37% ,respectively.

关 键 词：数字图像分析技术 棉花 氮素 营养诊断 地面覆盖度 叶面积指数 生物量 吸氮量 

分 类 号：S562[农业科学—作物学]