Study On Some Key Techniques For Extracting Vitality Index Of Pinus Massoniana Seedlings

The 18 th CPC National Congress clearly proposed to vigorously advance the construction of ecological civilization,strive to build beautiful China,and realize the sustainable development of the Chinese nation.Therefore,the accurate evaluation of seedling quality has become an important guarantee for promoting the construction of ecological civilization and modern forestry.At present,the evaluation index of seedling quality is mainly divided into morphological inde xes(seedling height,diameter and root conditions,etc.),physiological indexes(water content,chlorophyll content,nitrogen content,etc.)and dynamic indexes(growth rate,etc.).Traditional measurement methods are complicated to operate,time-consuming and laborious,and most of them are lossy detection.Therefore,it is of great significance to apply advanced technology to achieve the extraction of relevant vitality indexes of Pinus massoniana seedlings.This paper is for some key technologies study o f the vitality index extracting of Pinus massoniana seedlings.The three-dimensional reconstruction technology based on multi-vision and near infrared spectroscopy analysis technology were used to achieve the rapid,non-destructive and accurate extraction of some morphological and physiological indexes of Pinus massoniana seedlings.The main studying contents and achievements are as follows:1.This paper designed a set of morphological parameter extraction system for Pinus massoniana seedlings,the system based on multi-vision,non-contact technology.The system includes the establishment of seedling image acquisition hardware experiment platform and the design and implementation of software system based on Matlab software platform.The system uses the "Zhang Zhengyou calibration method" to calibrate the camera and obtains internal and external parameters.Then the image is preprocessed by the median filter and Laplacian edge sharpening algorithm.The spatial point cloud data of seedlings is obtained after image correction and stereo matching.And then obtained the spatial point cloud data of the seedlings by image correction and stereo matching,after that the three-dimensional model of Pinus massoniana seedlings was obtained by the point cloud splicing and Delaunay triangulation algorithm.Finally,the tree crown,tree trunk and tree root were auto matically identified through the differences in color information of each part of seedlings,the length and volume of each part are measured by correlation algorithm.After the evaluation of the experimental results,it was proved that the method fully meets the requirements of traditional forestry surveys and can replaces the traditional manual measurement methods.And the method will increase the accuracy and stability of the measured data and reduce the manual labor at the same time.2.Near infrared spectroscopy analysis technology is applied to predict physiological index(water content)of Pinus massoniana seedlings in this paper.The experiment uses the MPA series near-infrared spectrometer from BRUKER to measure the near-infrared spectroscopy information of Pinus massoniana seedlings.After preprocessing such as noise reduction and data normalization,the characteristic bands were selected by the backward interval partial least squares method and finally established the partial least-squares model to predict the relevant physiological index of Pinus massoniana seedlings,and evaluate the experimental results according to relevant evaluation indexes.It was proved that the partial least-squares algorithm based on near-infrared spectroscopy technique described in this paper is fully satisfied the requirements of actual measurement accuracy,it can be used as a substitute for complex traditional manual measurement methods.