Research On Key Technologies Of Damage Assessment System Based On Image Analysis

In the information warfare under the new situation,the damage assessment method based on image analysis will play an increasingly important role,but the method still has great difficulties in research and implementation.The research background of this paper comes from a national defense project.Through systematic research on relevant damage assessment technologies at home and abroad,the damage assessment process based on image analysis is developed,including image preprocessing,image registration,image change detection and damage effect evaluation.Several links.In the image preprocessing process,this paper focuses on the method of relative radiation correction,which can correct the radiation distortion between pre-damage and post-injury images.And the image preprocessing operation eliminates the interference of radiation distortion and noise on image processing.Pre-and post-harm images need to be registered prior to the image change detection process.This paper proposes an improved image registration algorithm based on SURF features and an image registration algorithm based on MSER-SURF features.The SURF feature has scale and rotation invariance.In this algorithm,the RANSAC algorithm based on nearest neighbor matching is introduced,which improves the accuracy of image registration.The MSER feature has robust affine invariance.For this reason,a method is proposed to transform the MSER feature into an elliptical shape and then normalize it into a circular region and then into a SURF feature point description.This method can be completed.Large angle deviation image registration.Aiming at the problem of multi-scale image fusion change detection,two different fusion algorithms are proposed in this paper: based on binary image fusion change detection algorithm and difference image fusion change detection algorithm.The binary image fusion change detection algorithm is used to obtain the change region in the multi-scale difference change image by the Otsu algorithm.The method has a faster processing speed but less precision.In order to solve the problem of insufficient precision,the difference image in the difference image is obtained by the MRF image fusion of the variable weight based on the difference image fusion change detection algorithm.Simulation results show that the algorithm is improved in accuracy.In addition,this paper proposes two different fusion strategies: single-pixel fusion strategy and MRF-based neighborhood system fusion strategy,which are applicable to different scenarios.There is no complete theoretical system for the damage assessment.By analyzing the relevant literature,this paper designs a damage assessment method based on the combination of geometric area change and texture feature change,and evaluates the damage by changing the area of the area and the change rate of the texture features.effect.In the evaluation process,the damage effect can be automatically calculated through the evaluation criteria in the expert knowledge system.Finally,the paper introduces the design of damage assessment system based on image analysis,and proves the feasibility of the research through simulation experiments.By summarizing the problems in the research process of the subject,the contents that need to be improved and studied are introduced.The research results of the thesis have been applied to the actual military projects,and the evaluation results are good.