Study On Chlorophyii Contents And Hpenotypic Paramenters Of Maize Leaf Based On The SVM And Regression Analysis

Maize is an important crop, and its planting area is the third largest in the world, second only to wheat and rice. Maize yield has important impact on the worldwide food security. Chlorophyll is important pigment for photosynthetic. It has an irreplaceable role in the growth of plants. Phenotypic traits are closely related to plant nutrition, biological and non-biological factors, and plant breeding.In this paper, we use multi-spectral camera to take images of maize leaf. Then the spectral information of maize leaf is obtained by using tool software; chlorophyll contents are detected by a chlorophyll mete. Based on support vector machine (SVM), we accurately classify three-period maize with the prediction accuracy, higher than 0.8. We establish a relational model between spectrum data and chlorophyll contents by using linear and nonlinear regression, and estimate chlorophyll contents under the experimental conditions. On the other hand, chlorophyll distribution of maize leaf is studied, and we find a distributing law of chlorophyll along the length and width direction of maize leaf. On the length direction, chlorophyll content in seeding stage is on the rise, relatively average in heading stage, and unstable in flower stage. On the width, chlorophyll content is basically symmetrical. In terms of phenotypic parameters of maize leaf, maize leaf contour curve fitting and mathematical expression of leaf shape are studied. And the experimental results using 25 sets of data show that maize leaf profile conforms to quadratic equation, correlation coefficients reaches 0.9 above.