Analysis Based On Remote Sensing Images Of The Cluster Bombs Airport Damage Effect

Timely, accurate assessment of the impact plays a very important role in thecampaign after decision-making.It can also accelerate the process and save war cost.inthe modern war, seizing control of the air war has become the key to theprocess.Therefore,as the Air Force is only rely on the airport, naturally the airportbecome the focus of attacking. Based on this,this paper uses digital image processingtechniques and data mining technology in airport damage analysis to study.As one of the issues in the field of pattern recognition, airdrome destructiondetection has important applications in the military field. In this paper, we make someresearches and progress on the detection of destruction of airdromes which were attackedby cluster-warhead. The images that we used are remote sensing images .Besides that, wehave done some further analysis to the destruction by methods of data mining.Since restricting of testing condition,a mathematical model of cluster-warhead ismade to simulate the result of attacking by cluster-warhead.In order to get a betterresult of recognization,gray transform is used to enhance the image.Before recognization,seed filling algorithm is used to extract airport runways mask image,so that we can getthe part that we interested in. In this paper, we proposed a neighborhood -based damagedetection algorithm referring to the gray characteristic of airdrome destruction bycluster-warhead in remote sensing images .Besides that, with combination of changedetection and this algorithm,we get a method to detect destruction on the dualimages.Then ,we use cluster algoritms to analysis the result of the last step.In thispaper, we have improved the genetic-algorithm-based k-medoids clustering algorithmand the k-means clustering algorithm,so that the program can determine the number ofclassifications automatically and we get a good result.