旱作马铃薯作物氮素营养快速诊断方法与机理研究

项目名称： 旱作马铃薯作物氮素营养快速诊断方法与机理研究

项目编号： No.31501219

项目类型： 青年科学基金项目

立项/批准年度： 2016

项目学科： 农业科学

项目作者： 孙红

作者单位： 中国农业大学

项目金额： 20万元

中文摘要： 马铃薯被确定为我国第四大主粮，对其全生长期进行营养诊断实施精细化管理具有重要意义。作为块茎作物，马铃薯冠层氮素营养快速诊断尚存在数据源干扰大、生长动态解析不深入和氮素诊断模型精度不高的问题，进一步开展“旱作马铃薯作物氮素营养快速诊断方法与机理研究”很有必要。为获取有效检测数据源，以近地多光谱成像、光谱辐射仪、超声波测距传感器和GPS多源传感器同步获取田间马铃薯作物表型参数（颜色、单位面积覆盖度、株高、光谱反射率、植被指数和位置信息），提出冠层多光谱图像精准分割和光学信号干扰消噪方法，消减土壤背景、杂草和开花等信息带来的干扰，解析不同生长期表型参数对氮素营养在地上和地下茎间 “峰式”转移的动态响应， 从而建立起适于马铃薯块茎生长阶段追踪的量化指标；根据多源检测参数与生长期马铃薯氮素之间的相关分析，优化氮素敏感检测参数，建立高精度马铃薯氮素营养诊断模型，最终实现旱作马铃薯作物氮素营养的快速诊断

中文关键词： 养分动态；多参数协同；光谱分析；图像处理；定量模型

英文摘要： The potato is one of the most important food crops, it is identified as China’s fourth major food. It’s important to improve the yield and reduce the chemical use. In order to satisfy the crop requirement of the field management time, position and reasonable fertilizer rate, the nitrogen diagnosis technique should be applied. The research will conducted on the method and mechanism of rapid diagnosis of potato crop canopy nitrogen content based on the preliminary study results. In order to obtain the effective data in the field, Multi-sensors will be used including multi-spectral camera, spectral radiometer, ultrasonic ranging sensor and GPS. The nitrogen content of potato leaf will be measured in the lab. The multi-phenotyping parameters will be extracted as color, unit cover rate, height, spectral reflectance, vegetation index and sampling location. Firstly, in order to reduce the data interference from soil background filed weed and potato flower, the multi parameters will be corrected with image accurate segmentation and spectrum denoising method. Secondly, the dynamic response of the sensor parameters will be analyzed with the growth stage and nitrogen nutrition changes. The relationship between multi-parameters, different growth stages and nitrogen content will be discussed. As a result, the quantitative indicators of growth stages and sensitive parameter of nitrogen content will be selected. Furthermore, they are used to establish the classification model of potato growth stages and nitrogen content of potato canopy. The key management period of potato (Tuber formation- growth stage) will be tracked, and the field nitrogen content distribution could be shown in GIS map based on the GPS data. The conclusion will be drown with the accurate crop needs and management space-time information. It will provide a rapid and nondestructive method for making precision fertilizer-decision in the potato field.

英文关键词： Nutrient dynamics;Multi parameter coordination; Spectral analysis;Image processing;Quantitative model