Inversion Of Chlorophyll Content Of Plant Based On Integration Of Geometry And Spectrum Phenotype

As one of the most important part of the earth's biologic chain,plant is the beginning of the earth's energy.Its growth is closely related to the production and life of the people.Therefore,how to efficiently and accurately obtain the growth and nutrition level of the vegetation and monitoring its various parameters has become the focus of study.With the development and application of hyperspectral remote sensing technology,it is more and more applied to various fields due to its non-destructive and fast characteristics.The applications of using vegetation hyperspectral remote sensing technology to monitor vegetation biochemical parameters are also widely applied.The development of three-dimensional laser point cloud technology makes the acquisition of three dimensional point clouds become faster and more efficient,which provides a technical method for obtaining an important parameter of plant:geometric parameter.In this paper,a self-developed portable crop growth analyzer is built and used to integrate the multi spectrum and laser point cloud data,and obtain plant's spectral and geometric information simultaneously,also,research about retrieving the chlorophyll content of the vegetation.In this paper,through the experiment of the selected broad-leaved plants,the spectral and point cloud data after integration are obtained.The inversion of chlorophyll using samples with different leaf angles is analyzed,and the effects of geometric information on the inversion of chlorophyll content are considered.The main contents are as follows:(1)The geometric information of the plant was extracted,including plant height,leaf inclination,and azimuth angle,which are compared to the measured parameters.It was found that the determining coefficient of the measured plant height and the value which is calculated by point cloud data is above 0.8,and basically conforms to the normal trend of 1:1;the determination coefficient of the calculated leaf inclination angle and the measured data is over 0.7,and the determination coefficient of calculated azimuth and measured data is over 0.9.(2)The spectral index model and partial least square method are used to retrieve the three scales of chlorophyll content:plants,leaves and part of the leaves.In all the vegetation indexes which are used in this experience,the effect of CIrededge and NDRE was better,with the coefficient of determination reached 0.794,0.5259 and 0.4358 on these three scales.The partial least squares regression shows better accuracy with vegetation index on plant scale,but the results on leaf and part of leaf scale are not as good as vegetation index.After the correction,the accuracy of chlorophyll content retrieval on these three scales has been improved relatively,and the corresponding RMSE value has decreased.(3)Based on the leaf inclination angle,the samples are divided into multiple angle intervals.And for each angle intervals,the inversion accuracy of the sample points are compared and analyzed by three scales:the plant,the leaves and the part of the leaves.It is found that the results of inversion accuracy are different in each angle intervals,and the difference are obvious.After spectrum angle correcting,the results of inversion accuracy are acting differently too.Through the comparison of the inversion of chlorophyll content in plants,leaves and part of leaves,this paper analyzes the effect of sample points on the inversion of chlorophyll content in different scales and different angles,and compares the inversion accuracy before and after the angle correction.The results build a foundation for the further study of the impact of geometric information in the inversion of chlorophyll content.