CAJ | 학술논문

选用扫描仪和无人机平台获取水稻叶片和冠层的数字图像,运用数字图像处理技术研究不同氮素营养水平水稻叶片和冠层的综合特征信息,从而应用于水稻的氮素营养诊断.结果表明:1)通过叶片叶绿素a含量和扫描叶片颜色参量之间的相关性分析,得到可用于诊断水稻氮素营养水平的叶片颜色特征参量B、b、b/(r+g)、b/r、b/g.通过叶片的颜色、形状综合特征信息与YIQ电视信号彩色坐标系统的参量建立氮素营养的识别模型,4个不同氮素水平的正确识别率分别为:N0(0 kg N·hm～(-2)) 74.9%,N1(60 kg N·hm～(-2)) 52%,N2(90 kg N·hm～(-2)) 84.7%,N3(120 kg N·hm～(-2)) 75%;2)无人机获取的田间冠层图像识别水稻氮素营养水平的综合特征参量是G、B、b、g、b/(r+g)、b/r、b/g、H、S、DGCI,选择相同的C_B参量建立冠层氮素营养的识别模型,4个不同氮素水平的正确识别率为:N0(0 kg N·hm～(-2)) 91.6%,N1(60 kg N·hm～(-2)) 70.83%,N2(90 kg N·hm～(-2)) 86.7%,N3(120 kg N·hm～(-2)) 95%.初步研究表明基于综合特征的氮素诊断模型区分效果比较好,利用叶片扫描图像和无人机识别与诊断田间水稻氮素是可行的.
선용소묘의화무인궤평태획취수도협편화관층적수자도상,운용수자도상처리기술연구불동담소영양수평수도협편화관층적종합특정신식,종이응용우수도적담소영양진단.결과표명:1)통과협편협록소a함량화소묘협편안색삼량지간적상관성분석,득도가용우진단수도담소영양수평적협편안색특정삼량B、b、b/(r+g)、b/r、b/g.통과협편적안색、형상종합특정신식여YIQ전시신호채색좌표계통적삼량건립담소영양적식별모형,4개불동담소수평적정학식별솔분별위:N0(0 kg N·hm～(-2)) 74.9%,N1(60 kg N·hm～(-2)) 52%,N2(90 kg N·hm～(-2)) 84.7%,N3(120 kg N·hm～(-2)) 75%;2)무인궤획취적전간관층도상식별수도담소영양수평적종합특정삼량시G、B、b、g、b/(r+g)、b/r、b/g、H、S、DGCI,선택상동적C_B삼량건립관층담소영양적식별모형,4개불동담소수평적정학식별솔위:N0(0 kg N·hm～(-2)) 91.6%,N1(60 kg N·hm～(-2)) 70.83%,N2(90 kg N·hm～(-2)) 86.7%,N3(120 kg N·hm～(-2)) 95%.초보연구표명기우종합특정적담소진단모형구분효과비교호,이용협편소묘도상화무인궤식별여진단전간수도담소시가행적.
The scanner and unmanned aerial vehicle (UAV) were adopted to take the leaf and canopy images, respectively. The images were corrected and applied to assess the rice nitrogen status on leaf level to canopy level by image analysis. The main results were as follows:1) According to the analysis of relationship between the content of chlorophyll a and the color parameters, the effective color parameters were discovered as B, b, b/(r+g), b/r, b/g. The regression was carried out on leaf level based on the distinctive characteristics in terms of color and shape. It was arranged in a particular form with C_B calculated at YIQ color system. The accuracy of the model under different nitrogen rates was as follows:N0(0 kg N·hm～(-2))74.9%, N1(60 kg N·hm～(-2)) 52%, N2(90 kg N·hm～(-2)) 84.7%, N3(120 kg N·hm～(-2)) 75%. 2) As far as the canopy level concerned, the synthesis characteristics were abstracted as G, B, b, g, b/(r+g), b/r, b/g, H, S, DGCI by calculating the relationship between the color parameters and nitrogen concentrations. The accuracy was as follows:N0(0 kg N·hm～(-2))91.6%,N1(60 kg N·hm～(-2)) 70.83%,N2(90 kg N·hm～(-2)) 86.7%,N3(120 kg N·hm～(-2)) 95%. The primary study indicated that the digital images taken from scanner and UAV could be applied to provide a cost-effective and accurate way to estimate rice nitrogen status.