Visual Analysis Method Research Based On Continuous Parallel Coordinates In Multi-attribute Data

Multi-dimensional multi-attribute data analysis and processing is one of the important contents of massive data analysis.In the field of oil and gas exploration,because of the low signal-to-noise ratio of the signal,the target characteristics in a single attribute are not obvious.Through the multi-attribute fusion analysis,it can highlight the geological characteristics and geological target characteristics.Based on this,this thesis studies the multi-attribute seismic data problem,and puts forward the multi-attribute seismic data analysis method based on visual analysis.The basic idea is to combine the visualization technology and human-computer interaction technology to make full use of the computer's processing ability and human subjective experience,on the one hand can avoid relying solely on the computer for data analysis of the accuracy of the problem,on the other hand can avoid relying on human-computer interaction to bring the complexity of the operation.It has a certain theoretical value and practical application value.This thesis study the visual analysis of multi-attribute seismic data.The main contributions are as follows:1.A multi-attribute data visual analysis method based on continuous parallel coordinates is proposed.Aiming at the visual analysis of multi-attribute data,a visual analysis process of multi-attribute data visualization,human-computer interaction feature extraction and fusion rendering is proposed.The first problem is the visualization of largescale multi-attribute data.In this thesis,the multi-dimensional multi-attribute visualization method based on continuous parallel coordinates is used to realize the display,extract and highlight the characteristics of multi-attribute data.Through the human-computer interaction process,the iterative analysis and extraction of the target feature are mapped.Based on this,the target feature is mapped into the transfer function of the fusion,and the target feature is presented in the three-dimensional space using the fusion technique.Human-computer interactions are through the entire data analysis process,real-time multi-attribute visual analysis is achieved.Through the simulation analysis,this method can effectively solve the problem of seismic data analysis.2.A multi-attribute data visual analysis method based on spatial information is proposed.In the multi-attribute data,the different attributes have certain correlation,and the target feature has correlation and continuity in space.Based on this,this thesis presents a multi-attribute data visual analysis method based on spatial information.The basic idea is to pick up the local information of geological objects through human-computer interaction,and use the spatial correlation and continuity of the target to highlight the spatial characteristics of the target.The basic process is to project the multi-attribute values in the scatter plot,characterize the spatial information of the voxels with the center of gravity and the spatial variance of the voxels projected to each pixel in the scatter plot,and classify the data according to the spatial information.Thus designing the transfer function to guide the fusion rendering results.This method can eliminate the interference of non-characteristic material and improve the feature extraction effect while preserving the characteristics of geological target.3.Design and implement a visual analysis system that integrates these two methods.The actual seismic multi-attribute data is used to simulate and provide practical verification for the method.This thesis focus on the analysis of mass multi-attribute data and put forward the visual analysis method.Through the simulation analysis,the method proposed in this thesis can effectively solve the visual analysis of multi-attribute seismic data.