Inversion And Verification Of Corn Leaf Area Index Based On Particle Swarm Neural Network Model

Leaf Area Index(LAI)is an important measure of crop growth.It is widely used in modern agriculture,forestry management,and other fields,and it has important academic research significance in the construction of terrestrial ecosystem and crop growth system model and agricultural ecological environment monitoring.In view of the load characteristics of GF autonomous satellites in China,the research and development of leaf area index inversion method can give full play to the potential of high score autonomous satellite technology application,and promote the application and development of GF autonomous satellites has important academic significance and value.In this study,Luancheng county,Huailai county and Langfang city in Hebei province were taken as the experimental bases.Based on vegetation index regression model,BP artificial neural network model,and particle swarm optimization neural network model.Firstly,GF remote sensing images of luancheng county were used to invert and verify leaf area index of maize at booting stage,then leaf area index of maize at whole growth stage in Huailai county,and finally leaf area index of maize at booting stage in Huailai county,Luancheng county and Langfang city.Mainly obtained the following research results:(1)The particle swarm optimization neural network model is adopted to invert the LAI of domestic GF satellites,which improves the inversion speed and efficiency,and has strong self-learning and adaptive ability,which can get the optimal solution in the convergence domain faster,and obtain the inversion results with high stability and good precision.(2)Summer maize is affected by light,water and growth rate in different growth periods,which leads to changes in leaf area index.Generally,during the whole growth period presents a process of slow growth,rapid growth and gradual decrease.According to this feature,the leaf area index inversion based on the particle swarm optimization neural network model is more consistent with the growth status of local maize than BP artificial neural network model,and 6 vegetation index regression model,which is conducive to the advance of maize yield estimation in the future.(3)In the three experimental area,the particle swarm optimization neural network model inversion than 6 vegetation index regression model and BP artificial neural network model inversion leaf area index has relatively high accuracy and stability,so has a very important role in the area,implements the inversion of leaf area index of regional characteristics.In conclusion,the particle swarm optimization neural network model inversion of leaf area index can meet the GF of satellite remote sensing image,region and more growth period,as a result,the particle swarm neural network model has great potential in inversion of leaf area index,has a certain universality,it plays an important value to improve the level of agricultural monitoring work.