Research On The Acquisition And Services Of Emergence Intelligence For Real-time Videos

With the rapid development of contemporary intelligent science and technology and the big data information industry,it has become a reality to obtain information bearing guiding significance for realistic emergency scenarios from large-scale real-time multimedia data resources.Meanwhile,it has become an important direction for the development of smart city information systems to jointly build an intelligent information service model which is engineered,intelligent,and collaborative though the combination of advanced information technology with emergency information management systems in contemporary cities.Real-time video,with extremely rich information content,is one of the most important data sources in the real-time data collection system of smart city.Therefore,how to obtain emergency information of different scenarios,different levels,and different types from real-time videos and apply it to advance warning,mid-event management and post-event conclusion of various emergencies has become a burning problem to be solved in the field of intelligence analysis and emergency management.Taking the acquisition of multi-level emergency intelligence content in real-time videos as the main logic of research,this thesis conducts a comprehensive study around three kinds of problems in the real-time video including analysis of multi-level emergency intelligence acquisition needs,technical exploration and service construction.First of all,the needs of emergency intelligence based on real-time videos in different scenarios are analyzed.Second,technology of video processing,target recognition and data mining are used to prove that intelligence resources of different levels can be obtained from real-time videos.Finally,the service mode and guarantee mechanism based on multi-level intelligence resources in real-time videos are explored,so as to comprehensively boost the development of emergency intelligence in the smart city.The main contribution of this study to the field of information science is demonstrated in the following three aspects.Firstly,the transformation path model of video data-oriented emergency intelligence resource is constructed,revealing the transformation mechanism and interaction relationship of intelligence resources in real-time videos,and expanding the theoretical basis of video resources as emergency intelligence services.Secondly,the technology method of obtaining the content information from multi-level video is put forward,which integrates the comprehensive application of video coding,image recognition,data mining and other interdisciplinary methods in the field of intelligence analysis,and the feasibility and applicability of information acquisition methods are verified through three sets of experimental methods.Thirdly,an emergency intelligence service model and guarantee system that integrates video resources are designed,exploring the law of using multimedia data sources to carry out intelligence services,and enriching the application prospects of video resources in emergency intelligence services.Focusing on the realization path,technical composition and situational application of obtaining multi-level emergency intelligence resources from real-time videos,three types of experiments are designed to verify the extraction of image intelligence,data intelligence and knowledge intelligence in real-time videos.The thesis contains 8 parts,and the specific content of each part is as follows:Chapter 1 Introduction.The background and significance of the research is introduced,and the research content,research methods and technical routes involved in this research are illustrated.Relevant domestic and foreign scientific research papers on emergency intelligence mining and services are comprehensively reviewed and analyzed.The system technology,mining methods and information services based on the real-time video are systematically reviewed,and their research are commented,trying to accurately grasp the shortcomings in the development and utilization of emergency intelligence of current real-time videos,so as to provide a reference for the research and innovation of this thesis.Chapter 2 Theoretical basis and related concepts.The content of this part explains the theoretical basis,related concepts and core terms involved in this research.The relevant theories guiding this research are investigated and discussed,and related concepts such as emergency intelligence system,emergency intelligence analysis,and emergency intelligence service are illustrated.The core key terms involved in the research such as image intelligence,data intelligence,and knowledge intelligence are defined.Chapter 3 Emergency intelligence demands based on real-time videos.This part is the logical starting point for developing emergency intelligence services based on real-time videos.Firstly,the innovative demands of developing emergency intelligence services based on real-time monitoring videos in the smart city are analyzed.Secondly,the demands for expanding intelligence analysis methods using real-time videos as the data source are analyzed.Moreover,the construction demands of emergency intelligence service system are analyzed from the perspective of process planning and stage response.Finally,the content demands of acquiring multi-level emergency intelligence based on real-time videos are analyzed.Chapter 4 Extraction of emergency image intelligence from real-time videos.This part of the research is the first step of video intelligence processing technology.The extraction path of image information are studied and the architecture,development and management of image information extraction are discussed in terms of three aspects including content,characteristics and processes of the smart city-level large-scale emergency video resources.After demonstrating the principles of codec and real-time compression technologies for emergency video processing,the feasibility and applicability of extracting image information from emergency video are verified through experimental design.Chapter 5 Extraction of structured data intelligence from real-time videos.This part of the research is the second step of video intelligence processing technology.First,the extraction path of data intelligence from the three aspects of content,characteristics and process of emergency images is studied.Secondly,the target recognition method is used for emergency images to achieve the transformation from target image resources to structured data resources,so as to explore deeper levels of emergency intelligence content from the level of structured data.Finally,the technical exploration of the identification and transformation of key targets in emergency images into structured data are completed through the experiment design,and the feasibility and applicability of data intelligence extraction in emergency intelligence scenarios are verified.Chapter 6 Acquisition of decision-making knowledge intelligence from real-time videos.This part of the research is the third step of video intelligence processing technology.First,the acquisition path of knowledge intelligence is studied from three aspects of content,characteristics,and processes of structured data identified in the emergency video.Second,the composition framework and development management of knowledge intelligence derived from the structured data are analyzed theoretically.Finally,the data extracted from the emergency video is verified through experiments to be used to effectively obtain various types of knowledge intelligence resources with guiding significance by data mining and other intelligence analysis methods,and the feasibility and applicability of acquiring knowledge intelligence based on emergency data are explored.Chapter 7 Service framework of emergency intelligence integrating video resources.This part is the ending point of the research that an emergency intelligence service model that integrates video mining is constructed by integrating all levels of intelligence resources obtained from real-time videos with other emergency intelligence resources,the construction methods of intelligence service resources based on real-time videos are analyzed,and an information service guarantee system based on real-time videos is explored.The study believes that it is necessary to form a construction mechanism for all kinds of emergency information resources obtained from real-time videos in the smart cities from three aspects including content classification and integration,resource integration and management,as well as comprehensive quality assessment.The intelligence analysis methods obtained from various emergency information in real-time videos should be organically combined to build a coordinated and linked emergency intelligence service model that can comprehensively respond to risks.At the same time,in the national macro-governance framework,a clear guarantee level,comprehensive guarantee capabilities,and sound legal protection need to be established for the emergency information service guarantee system that integrates video mining.Chapter 8 Summary and prospect.First,the drawbacks of this research in terms of research methods,experimental design and application analysis are summarized.Then based on the deepening of the contemporary information science and technology revolution and the historical background of the rapid development of China’s information industry,the emergency intelligence research based on real-time videos in the future is prospected.