Research On High-efficiency Rendering Of 3D Video GIS

With the rapid development of social economy,the Chinese government continues to enhance the level of urban public facilities in recent years.Currently,cameras and video surveillance systems,which are scattered in every city,play an important role in maintaining public security and urban traffic safety.Tens of millions of data can be collected by a camera in a single day.Also,as the development of camera hardware technologies,some cameras with high precision have emerged to make it possible to collect large amounts of video data with detailed texture information.So,how to make full use of those data is a field worthy of further explore.As it is known to all,data is the blood of GIS.And GIS serves as a platform for various spatial and attribute data.Being widely used in all field,GIS is being better understood among humans.3D GIS is intuitive and informative so that commercial 3D digital earth software,represented by Google Earth,is popular with the public.Thus,it is valuable in theory application and practical application to combine multimedia data,such as video data,and 3D GIS,which is a method to take advantage of real-time and dynamic feature of video data and abundant geographical environment of 3D GIS.However,the combination of video data and 3D GIS consumes a great number of computing resources and requires an excellent concurrency control capability.This has become a huge stumbling block in the development of 3D video GIS.In order to better combining video data and 3D GIS and ensuring stability and fluency of the combination,it is essential to optimize sections which are over-consumption of resources.That means optimizing the process of rendering 3D GIS with video data.This paper studies the efficient render of 3D video GIS.The main research contents of this paper are as follows:(1)The process of rendering 3D GIS with video data is rendering 3D Scenes with images extracted from video data.So,the efficiency of extracting images from video data,which means generating textures applicable for 3D Scene Rendering,is one of the critical factors affecting the efficiency of rendering.This paper analyzes the encoding principles of digital video and some basic concepts,such as representation methods of digital video,video models and popular video formats.The principles and methods of extracting image from video data are discussed in this paper.Also,texture mapping and the process of rendering 3D scenes with texture images are explored.These works lay a good foundation for the next step.(2)It consumes a significant amount of computing and 10 resources to operate the 3D scenes system,convert video data to image and respond to user actions.In particular,when it comes to the conversion of video data under large-scale system,multiple video's simultaneous display in 3D scenes inevitably leads to system blocking,lagging of video or even system crashes.So,in order to allocate system resources optimally,it is essential to use multi-threading technology to classify computing tasks into independent types,such as the processing of a single video and responses to user actions.Also,communication and collaboration between threads should be ensured.When considering multithreaded programs' control on the number of threads,a single thread's occupation of system resources and the system overhead for thread switches,it needs to keep the number of threads as small as possible which means that videos active simultaneously should be as few as possible.According to the location information in video data and the parameter in active window,this paper test the visibility of video data by utilizing grid to segment experimental scenes.In this process,invisible videos will be put into an inactive state and the resolution of videos will be analyzed to release resources occupied by indistinguishable video data,which help to adjust the system resources.(3)3D video GIS systems have practical application value.They often involve the use of multiple techniques and the application of software to implement.A 3D video GIS system is implemented to be a framework for experiments in this paper.Its render efficiency is detected on the basis of general indexes.Moreover,the experimental results are summarized,compared and analyzed in this paper.