Analysis On Hyperspectral Characteristics Of Rice Under Different Nitrogen Levels

Nitrogen is an important nutrient necessary for crop growth,and plays a key role inrice regulation.Hyperspectral remote sensing is an important mean for estimatingbiochemical components such as nitrogen pigment content and parameters such as leafarea index（LAI） and biomass in crop plants. and SPAD. The correlation between thepigment concentrations,leaf area indices, biomass above ground and the red edgeparameters of rice were analyzed in this experiment. In this study, field experimentspotted and with different rice varieties,and hyperspectral reflectances of the canopy, theleaf on the main stem of two varieties of rice are measured by a SVC HR-1024in fieldand indoor under5nitrogen support levels in different positions.The correlation betweenspectra and contents of chlorophyll, carotenoid of canopies and leaves were showed.The results showed that，different cultivars and different leaf position showeddifferent canopy spectral reflectance and derivate spectra,especially in visible band andinfrared band,reflectance at visible region increasing with the increased of nitrogen level.Considered their practicality, the sensitive bands for canopy were520nm-565nm，580nm-685nm，745nm-905nm.The leaf area indices (LAI),above ground dry biomass(ADM) and dry leaf mass(DLM) were very significantly correlative to the spectralvariables (?) and (?).The pigment contents of canopies and leaves were alsocorrelative to the spectral variables (?) and (?).This study puts forward an effective way of hyperspectral remote sensing datacollection method in the field, preliminary proved the spectral characteristics of fullheading stage of rices, analysised and compared the spectral features, and estimation ofthe red edge parameters of rice were analyzed. This study provides a theoretical basisand technical support for the extracting rices remote sensing data information, nutritiondiagnosis and rices production and management of informationization.