Preliminary Studies Of Artificial Spectral Characteristic

The 70% of the land surface is covered with vegetation. Plant species diversity and plant community complexity directly impact on remote sensing images interpretation. Usually, those in remote sensing images are decided by the spectrum of plant community and spectrum of soil. The effects of the weather condition on the remote sensing images are very difficult for interpreting. With the development of remote sensing technology, high remote sensing spectrum, it plays a good role in against the vegetation information and classification of vegetation and vegetation features identification.The ASD portable holding spectrograph is used in this study. With corn, flue-cured tobacco and greening vegetation as study objects, higher spectral test is done to research the plant spectrum characteristics. In the analyzing process, the original spectrum and the transform spectrum are used to do correlation analysis, regression analysis and red edge characteristics analysis of vegetation nitrogen content and chlorophyll content, and then to find the characteristic wavelengths which nitrogen and chlorophyll effect.The main results are as follows:1. By the study on the canopy spectrum of corn with different growth period and nutrition levels, the regression fitting result of the ratio between the canopy spectrum and the broad band is relatively good in the tasseling stage of corn, and the fitting accuracy of the third order polynomial tends to 0.9628; During the whole growth period, the difference of red edge slope of corn canopy spectrum is obvious. The amplitude of red edge slope is in the range of 0.02~0.27, and the site of red edge is in the range of 680nm~780nm; The correlation between phosphorus and corn spectrum is not significant; The red edge slope is lower when adding proper magnesium only in the tasseling stage and dough stage.2. During the plants spectrum transforms, the first order differential transform can make the difference of non-sensitive wavelength decrease and make the difference of sensitive wavelength increase. The second order differential transform mainly increase the difference of green wavelength and red wavelength, the effects on the other band of wavelengths are not significant; The canopy spectrum difference of flue-cured tobacco is not obvious under different fertilizer levels. With first order differential transform, the greatest value of red edge slope of flue-cured tobacco appears at 720nn and 715nm; During the correlation analysis of flue-cured tobacco canopy spectrum, the fitting accuracy of quadratic polynomial is highest and the fitting accuracy of cubic polynomial regression analysis is highest in the near-infrared wavelength.3. From the spectrum test of some greening vegetation in the campus, we can see that the shape of plants spectrum curve is basically consistent except of local with greater volatility. The spectrum reflectance is higher in the green wavelength but lower in the red wavelength ; In this study, the spectrum reflectance of all plants is lower than 0.05 in the wavelength between 350nm~500nm, but, it appears some laws in the near infrared wavelength: The spectrum reflectance of trees is mainly in 0.4~0.5; The spectrum reflectance of shrubs is in 0.2~0.7; The spectrum reflectance of herb with a big span is in 0.25~0.75; With the impact of seasons, the spectrum of different plants are different in various seasons.