Remote Sensing Monitoring Of Unlicensed Mines

Mineral resources are valuable non-renewable resources of human, so thesustainable development of mineral resources is a set policy of our country. It isimperative to search and crackdown on illegal mining.Using remote sensing imagesand take a set of automated target figure spot extraction method for unlicensed minesextraction has great significance.due to most of the extraction of unlicensed mineadopt remains pure visual artificial extraction method, so the purpose of this article isto seek the automated extraction of unlicensed mine map spot and conducting theresearch.Tish Paper in the north of a small piece of Jiujiang as the experimental area andusing the SPOT-5images.At first, make the necessary image preprocessing.Then,select the object condition for establishing the characteristics of the function and themembership function, take object-oriented classification for processed image, thenaccess all types of terrain map spot.Reuse of image difference method for two imagechange detection, detection of class graph of spot.After the above,stake the two layersof mine map spot and the data, through the analysis of GIS platform for suspectedunlicensed mining spot.Finally,validation this spot by visual interpretation andverified in the field.In this kind of object-oriented classification technology, first of all,using rationmethod to get the more suitable segementation scale for this experimental plot areaand get a more appropriate segmentation scale, determine the appropriateclassification level.Then get all kinds of feature samples, the selection on theirspectral and texture features properties to determine their condition and membershipfunctions, and then classified by all kinds of features of map spot, and finally theclassification results made a simple evaluation precision.In change detectiontechnology, using the difference method for analycised image, to calculate the band ofevery class on the two images of gray scale. Then,get suitable differential bands andgenerated binary diagram of differential operation, then scatter of binary chart simpleclassification, Finally,get variation.From the point of this article research results, the way of the object-orientedclassification in combination with change detection has certain feasibility for theextraction of target chart spot.And it’s has a profound significance for the automaticextraction of target chart spot research.