Identification And Characterization Of The Highlights Of The Sunspot Penumbra Fiber

Digital image processing is a process of using a computer to deal,which concerts image information into digital information.Digital image processing mainly use some technology,such as noise reduction,image restoration,image segmentation and feature extraction and so on,to improve the quality of the original image and make original image clearer,so that achieve the goal of conservation and research.At present,with the continuous development of technology,the development direction of digital image processing toward higher and deeper.That provides important technical foundation for astronomy.The sun has drastic change all the time,the activities of sunspot in the photosphere is one of the typical solar activities.The sunspot is made up of umbra and penumbra,and penumbra usually includes bright filaments and dark fibrils.Research shows that there is a huge magnetic field in the sunspot,so the magnetic field also exists in above the sunspot penumbra.And the form of the fibrous penumbra is related to the existence of the magnetic field.At the same time,the study found that the penumbra connect with the form of the Evershed flow and flow of the rising hot gas.So in order to better study the relationship of the formation of the penumbra and the magnetic field,and the Evershed flow formation,we should identify penumbra and study the properties of the penumbra.Therefore,it is important to identify and study the penumbra.Through the observation discover the shapes of bright filaments appear to comets,which has a bright grain close to the umbra,and following is filamentous.So base on the that feature we can think the bright filaments is the representative of the bright penumbra,so identifying the bright grain in the bright filaments and analysis of relevant features can help we to better understand the features of the bright filaments.At present,the main method of identification is band on high pass filtering.Essentially,this method is choosing to suitable threshold value to extract the bright grains.But the process of choosing suitable threshold is very complex.At the time,the result is good or bad should through observing the judgment,which has subjectivity.So in order to solve this problem,we propose a new identify method,which is based on Morphological Reconstruction.The specific process is as follows:firstly,using the linear filtering technology to reduce the noise of the images;then,extracting the area of penumbra,which need extract the frontier of the umbra and the sunspots;at last,obtaining the reconstruction image by using the Morphological Reconstruction,and obtaining the diffidence image by the reconstruction and original image,than extracting the bright grains by using the morphological operation,such as normalization and binarization method.In this paper,in order to better display the identification process,we choose the he first frame of New Solar Telescope(NST)sequence diagram as representative to show.And in order to verify the feasibility and universality of the proposed method,there are two groups of contrast test.One is to identity two difference data and obtain the result images,we will from the integral result to prove the universality of the proposed method.The two difference data respectively are New Vaccum Sloar Telescope(NVST)and the Solar Optical Telescope(SOT).Other contrast test is to obtain the result image of simulating the high pass filtering method,and compare with result image in this paper.Through this method can prove the feasibility of proposed method.At the same time,in order to prove the validity of proposed method,we chose the subjective evaluation and objective evaluation to prove.As far as the subjective evaluation,we randomly selected five data from three kinds(there are SOT,NVST and NST),and then we use the method of proposed to identify those datum.And last,we use artificial tag statistics to count the number of bright grains,the correct number and the errors number.And calculation accuracy and error probability to evaluate the method of proposed in this paper.As far as objective evaluation,we use the same data with existing literature.The data is the Solar Optical Telescope(SOT)sequence diagram(sum to 764 images).And we count the mean of area and intensity to compare with the result of exist literature.We are from the aspects of objective to evaluate the validity of the proposed algorithm in this way.