Acquiring And Analyzing The Optical Properties Of Maize (Zea Mays) Leaves

One of major part in precision agriculture is the acquisition information of plant growth and the diagnosis of the growth condition applied by remote sensing technology and computer graphics. In addition, water or nutrient absorption on plants is closely related to leaf structure, biomass, pigment etc., leaf is an indicative organ for plant photosynthesis. Measurement of optical properties varied with external conditions, such as nutrient use efficiency, adaption to environmental stress, is of promising prospect. Therefore, characterization of leaf optical properties, including accurate and efficient measurement of indices of maize (Zea mays) leaves and the quantitative analysis methods meets the needs in remote sensing and application of crops. In this thesis, using maize (Zea mays) leaves as research object, optical principle and model were applied to optimize the experiment conditions of the self-developed optical acquisition system. Background elimination for maize (Zea mays) leaves based on the BPLT model was established, experimental error was reduced by the correction of the distribution of Spectralon. Moreover, light reflectance distribution compared with multi-angle of nitrogen stress or water stress measurements in maize leaves was completed. The main contents and brief conclusion are as follows:(1) In order to precisely acquire leaf reflectance spectra, influence of background on leaf reflectance spectra was studied. Experiment was conducted to discriminate the characteristics of 60 maize (Zeamays) leaves based on 8 background materials and leaf chlorophyll concentration. BPLT (Background Plate) model, which premised on a set of assumptions,R=R12+(1-R12)*(1-R21)*t2*(1+R221*t2)*R21, was promoted and applied to remove the influence of background material. To verify this model, Analysis of Variance (ANOVA) was conducted. The results indicated that maize (Zea mays) leaf could be effectively estimated by the means of BPLT model with determined coefficients (DC) greater than 0.68 and residual sum of squares (SSE) less than 7.5. Meanwhile, leaf chlorophyll concentration at high levels was stable with the T-K-B background. As with the ANOVA, vegetation indices NDI, NDVI, SAVI, OSAVI, ARVI were better by One-way ANOVA while vegetation indices PSSR was better to study the influence of different chlorophyll concentration.(2) The preference of the homogeneity of the reference Spectralon panel was required to be calibrated to eliminate the corresponding factor. Bidirectional reflectance distribution factor (BRDF,fr), directional hemispherical reflectance factor (DHRF, p) as well as the measured data was held on to the conduction of fr,leaf. It was considered sensor zenith angle θr,degree=30°, of which was the lowest average root sum square (Rss=0.046), as the standard angle to decrease the error combined with fitting. Coefficient factor (Y) was plotted as a function of θr (in radiance, X) of different illumination angles Y=1.3938cosX-0.2127 (θi=0°～45°). The ultimate mathematical expression of fr,leaf was Thus, the error suffered from the curvature of the optical fiber or redundancy of repeated measurement data was reduced.(3) Light reflectance distribution compared with multi-angle of nitrogen stress measurements on maize (Zea mays) leaves was conducted based on the above study. Matlab GUI interface was developed for batch processing of optical data of maize (Zea mays) leaves. It was divided into four parts:data importing, data retention, calculation of fr,leaf and figure. Concerning different wavelength, different zenith/azimuth angle of nitrogen stress on maize (Zea mays) leaves, polar plot added with ANIX (Anistrophy Index) was demonstrated. Results showed that the maximum reflectance peak of maize (Zea mays) leaves was out of the principal plane besides the specular direction within increasing illumination zenith angle, in that case, the ideal experiment condition could be at θi=45° while wavelength of 680nm was more adaptable to analyze the characteristic of its distributing law. Moreover, the middle part of maize (Zea mays) leaves interacted with different illumination zenith angles were better to compare the optical properties among themselves.