Phenotypic Characteristics Of Fagopyrum Esculentum Based On 3D Reconstruction And Visualization

China is the birthplace of buckwheat crops,with a wide range of cultivation.Fagopyrum esculentum is mainly planted in Shanxi Province.Because Fagopyrum esculentum has high nutritional value,drought and cold tolerance and short growth cycle,it is of great significance to study the phenotypic traits of Fagopyrum esculentum plants for the development of modern agricultural breeding and seed selection technology.Traditional methods of measuring phenotypic traits by hand have some problems,such as strong subjectivity,high labor intensity,data damage and so on.In recent years,computer technology has been widely used in the field of crop phenotype.The information stored in 3D point cloud is more than that stored in 2D image,which can extract more feature information to analyze plants.At present,three-dimensional reconstruction technology and crop phenotype mainly focus on corn,soybean and rice plants,but less on common buckwheat plants.Therefore,it is of practical significance to provide a rapid and nondestructive method to measure the phenotypic traits of common buckwheat plants.In this paper,the phenotypic traits of Fagopyrum esculentum were measured and analyzed by using the three-dimensional reconstruction model of Fagopyrum esculentum.In this paper,the dense point cloud of Fagopyrum esculentum potted plants based on motion recovery structure and multi-view stereo vision is used to analyze the characteristics,and the voxel filter,conditional filter and statistical filter are combined to remove the noise points caused by human or environmental factors.And regression analysis is carried out with the data measured manually.The main innovations are as follows:(1)A three-dimensional reconstruction model of Fagopyrum esculentum based on surface motion restoration is proposed.The reconstructed model is preprocessed.(2)The paper uses a combined method of voxel filtering,statistical filtering and conditional filtering to remove the filter for the point cloud already acquired after the reconstruction of the sweet buckwheat plants.(3)Taking Fagopyrum esculentum potted plants as the research object,the algorithm for calculating phenotypic traits such as plant height,leaf area,leaf angle and leaf number of Fagopyrum esculentum plants was proposed based on the point cloud data obtained from surface motion restoration.The results show that the reconstructed model proposed in this paper has high adaptability to the measurement of different phenotypic traits.The determination coefficients of plant height,leaf area,leaf angle and leaf number measured by three-dimensional measurement and manual measurement are higher than 94%.The measurement algorithm used in this paper has high accuracy.The study on phenotypic traits of common buckwheat provides the most suitable planting scheme for growers,so as to improve the yield of common buckwheat,better adapt to market demand and realize its economic value.