Study On The Standardization Of MOST Series Massive Image Data Sets

The brain,as the main part of the central nervous system,is one of the most complex and most important organs of the organism.Its structure and function mechanism is a hot and difficult point in the field of brain science.In recent years,with the development of neuron labeling technology and imaging technology,especially the appearance of whole brain range and submicron high resolution optical microscopy provides a powerful tool for the research of brain science.However,because of the diversity of research purposes,the diversity of dyeing methods and the difference of imaging systems,a large,complex,diverse and different quality data set has been produced in the process of brain science,which has brought difficulties to data analysis and application.Therefore,in the process of data production,how to solve the difficulties in data analysis and application and the standardization of mass image data has become an important problem.The basic requirements for standardized data are the integrity of data and related information,the accuracy of stitching and registration,the consistency of the global gray scale,the unity of the anatomic coordinate system,the storage format and the naming rules.To obtain the standardized data,improve the image quality,make the obtained data meet the application requirements,the user can analyze and apply it directly.This paper studies the standardization of the MOST series of massive image data sets,analyzes the characteristics of the data in an all-round way,and is based on the standard of the massive MOST series image data set.This research consists of the following three parts:(1)To obtain standardized data sets and standardize the process of image processing,an image mosaic method based on interrelated information is proposed to realize the automatic splicing of panoramic images.The fusion of multi-channel data is realized by the method of phase correlation,and the correction algorithm and image based on the image gray compensation are studied.The enhancement algorithm realizes the global consistency of the gray level of the data set,realizes the standardization of the data set coordinate system through coordinate,two-dimensional alignment and three-dimensional registration,and standardization of the derived data.(2)To quickly and effectively evaluate the data set after the standardization process,the subjective and objective are established.The integrated evaluation system can comprehensively evaluate data integrity,global gray consistency and data quality.(3)To realize the process management in the large-scale data production,the data file system is managed and the automatic fast checkout method is developed to ensure the integrity of the data set and avoid the data from the imaging system.In the process of transmission and calculation of the storage center,it is lost in the process of transmission and calculation,and the information management system is used to manage data set and data related information in real time in order to realize the production scheduling of data.In this paper,from data processing to data evaluation to process management,the standardized processing flow is built and automatic and applicable image processing methods are developed to minimize artificial participation in the process of mass data processing and avoid errors caused by human factors.The evaluation method combining subjective and objective is used to evaluate the processed data quickly,to ensure the quality of the data and to improve the reliability of the image processing algorithm.The management of data processing flow can accumulate standardized data for brain science research,and provide important resources for brain science researchers to obtain brain science data quickly and conveniently,and provide the basis and guarantee for scientific sharing and management of brain science data.