Font Size: a A A

Research On Shape Reconstruction And Size Measurement Of Irregular Particle

Posted on:2024-01-24Degree:MasterType:Thesis
Country:ChinaCandidate:W B FanFull Text:PDF
GTID:2530307127960869Subject:Computer technology
Abstract/Summary:PDF Full Text Request
Clouds are polymers suspended in the air after the liquefaction or condensation of water vapor in the atmosphere.In order to understand the evolution of cloud droplets and irregular ice particles,it is necessary to study the cloud microphysical characteristics.Interferometric Particle Imaging(IPI)technology is a measurement method based on particle scattering light distribution.It has the characteristics of high measurement accuracy,large sampling volume,fast processing speed,etc.It is widely used for the measurement of size,position,phase and other parameters of spherical particle.In this paper,the principle of interferometric imaging of irregular particles is analyzed,and the shape reconstruction of irregular particles is realized by combining IPI technology and deep learning method.The algorithm based on image processing is used to realize the accurate measurement of particle size,which is verified by the IPI experimental system.The main work of this paper is as follows:1.Based on the transfer matrix theory,the interference imaging theory of irregular particles is analyzed;The real shape of particles is obtained through microscope observation,and the distribution of emitting points is determined.The simulated interference out-of-focus images of irregular particles are simulated by mathematical model;An IPI system is built to collect experimental interference out-of-focus images of irregular particles.2.The method based on deep learning is used to solve the problem of particle shape reconstruction.The data set for network training is prepared.Dense Unet network model is introduced.MSLE loss function and NPCC loss function are used for network training respectively,and their reconstruction performance is compared;The advantages of Dense Unet are verified by comparing other networks used for image reconstruction..3.On the basis of particle shape reconstruction,the particle size is measured.The image processing algorithm is used to calibrate the object in the image,calculate the pixel ruler,and convert the pixel information into the actual length;The size and shape characteristics of particles are analyzed respectively,and the feasibility of the method is verified by calculating the average relative error between the reconstructed image and the real image.
Keywords/Search Tags:Interferometric particle imaging(IPI), Irregular particle, Deep learning, The shape of particles, The size of particles
PDF Full Text Request
Related items