The Design And Construction Of Remote Sensing Image Targets Interpretation And Training Platform

Remote sensing image intelligence is the core element of geospatial intelligence,and remote sensing image interpretation is the most basic way to obtain the intelligence of remote sensing image effectively.Research on the integrated integration training of multivariate,multi-scale and multi-resolution ground and space targets and interpreting the valuable remote sensing image information accurately and efficiently by relevant business personnel are of theoretical and practical significance to fast acquisition of Geospatial Intelligence.Focusing on the key technologies and methods of interpretation training about remote sensing image targets,this paper summarizes the image features of remote sensing images and form the uniform training rules by analyzing the types and characteristics of various image interpretation targets.Combined with the HTML5 technology,the platform of remote sensing image targets integration training is constructed,which can fully meet the training requirements of the remote sensing image target interpretation.The main work and innovation of this paper are as follows:1.It summarizes the remote sensing image interpretation characteristics of typical ground targets,and extends the interpretation target from the ground to the deep space.Based on the geometric,orbit and attitude characteristics of the spatial target,the static and dynamic interpretation features of the remote sensing image are analyzed,and the corresponding basic knowledge base is established.This paper summarizes the feature-based remote sensing image interpretation method,the multi-temporal remote sensing image contrast analysis method and the three-dimensional contrast interpretation method,which provides the theoretical basis for the unified space-air-ground integration target interpretation training.2.A cognitive training model based on feature of multi-time and multi-dimension spatial remote sensing image is proposed,which provides a whole cognitive process for the training of remote sensing image interpretation,and reveals the internal relations and rules among the interpretation characteristics of the geographical spatial objects.By comparing the characteristics of different phase remote sensing images and the correlation analysis of the two and three dimensions of the image object,the fast and efficient training for typical targets of remote sensing images is realized.3.The HTML5 and WebGL technology are introduced into construction and development of the platform and fuction of three-dimensional visualization based on WEB is the research emphasis.It solves problems existing in interpretation and training platforms of traditional standalone remote sensing image,such as single function,high dependence on plug-in,weak system integration,complex operation,and greatly improves the operability and visibility of interpretation and training.4.In this paper,the improved Analytic Hierarchy Process model is applied to remote sensing image interpretation training.The AHP evaluation method based on dynamic weight of error rate about interpretation training is proposed.Through the different error rates,are determined the dynamic weight of training personnel.Thus,the weighted score based on error rate is obtained and compared with the general quantitative results,which makes a comprehensive and objective evaluation.The experimental results show that the evaluation model is correct and applicable.5.An integrated training platform of remote sensing image targets is designed,which provides a new web training method and way for the image information interpretation,in this paper.