Font Size: a A A

Research On The Feature Extraction Method Of The Bright Spot Of The Sun Photosphere Image And The Chromosphere Image

Posted on:2019-05-02Degree:MasterType:Thesis
Country:ChinaCandidate:J P XiongFull Text:PDF
GTID:2438330566983730Subject:Software engineering
Abstract/Summary:PDF Full Text Request
Photospheric bright points and chromospheric bright points are the smallest scale magnetic structure,and they are considered to be a tracer of the thin flux tubes.It is of great significance to study photospheric bright points and chromospheric bright points to promote the study of the expand of the thin flux tube with the increasing atmospheric height,and the coronal heating.At present,scholars have a good understanding of the characteristics of the photospheric bright points.The traditional methods used for photospheric bright points typically employ a track-after-detect strategy.Due to the track-after-detect strategy does not consider the changes of some characteristics with the evolution and inevitably misdetects some bright points that are weak or adhere to granules,then causes long-lived bright points to be segmented into several short-lived discontinuous sections,and the consequence is that the evolution processes of bright points might be incomplete.A new algorithm had been proposed to identify and track solar photospheric bright points by using a track-while-detect strategy in the three-dimensional data cube consists of position,time axis and intensity,this track-while-detect method has overcome the shortcomings of the track-after-detect strategy.Compared to the photospheric bright points,the chromospheric bright points is less studied,The reason is that the edge structure of the chromospheric bright points on the chromosphere image is not obvious enough,and the intensity is close to the surrounding structure,which brings great difficulties to extract it.Most of the methods for identifying of chromospheric bright points are by manual.In addition,the method of extracting corresponding bright points in photosphere chromosphere is also manually matched.The manual method is obviously not practical with the increasing datas.Therefore,in order to study more quickly and effectively,it is necessary to apply image processing method,numerical analysis method,etc.,to process and study the highlights of photospheric bright points and chromospheric bright points.In view of the above problems,in this paper,a method is proposed to automatically correspond to the photospheric bright points and chromospheric bright points,which consists of three parts.At first,a image registration algorithm reference to Harris feature point detection algorithm is designed to eliminate dislocation between photospheric and the corresponding chromospheric images resulted from the collecting method and telescope.Then,identifying the bright points in the images after registration,for photospheric bright points,the identifying and tracking method is still adopted the track-while-detect method that proposed in 2015,for chromospheric bright points,the algorithm is also reference the track-while-detect strategy,we identify and track them in a four dimensional consists of the position,time and intensity by using region growing method and fully consider the changes of characteristics with the evolution.Finally,we propose an algorithm to correspond the photospheric bright points and chromospheric bright points automatically,and the make analysis of characteristics for the corresponding bright points.The experimental results show that the analysis of the corresponding bright points' characteristics are similar with other scholars who extract bright points in manual methods,it illustrates that the methods we proposed to correspond bright points in photosphere and chromosphere are valid and accurate,and the method provides an automatic way to study bright points in photosphere and chromosphere easily,which has changed the traditional manual way.The identifying and tracking method for the chromospheric bright points we proposed in this paper can extract the bright points on chromospheric images completely and effectively,and then provides a better automatic identifying and tracking method for chromospheric bright points.After In addition,the method we proposed in this paper can also be used to study the other wavebands.Thus,the study of thin flux tubes extending and the study of solar magnetic field will be better promoted.
Keywords/Search Tags:Photospheric bright points, Chromospheric bright points, Track-while-detect, Image registration, Corresponding of bright points
PDF Full Text Request
Related items