Feature Extraction Of Mine Surface Elements And Land Use Dynamic Monitoring By Remotely Sensed Data

In the development of human society and construction of national economy,mineral resources are the indispensable material resources. As limited resources, mineral resources are not renewable, for the formation which a long time. So the sustainable development is established in the orientation of mineral resources development in our country. Since the 20 th century, the exploitation of mineral resources got a rapid development, which also caused many ecological environment problems. Therefore, the development and utilization of mineral resources situation get more and more attention. Feature extraction in the mine surface element, which a series of extraction methods based on remote sensing images, have great significance on the development and utilization of the mineral survey. Even the artificial visual interpretation is the common method with extraction of feature distribution using high resolution images, but it consumes a large amount of work force. Therefore, it is of great importance for providing the scientific basis for real-time monitoring over mining situation, as well as for conservation and optimal management of the environment of the mine development, that researching the feature extraction of mine surface elements and dynamic monitoring over land use.Contents of this article are part of the subprojects for the National Natural Science Foundation “The study of method on extracting and automatic identification of the multi-factor hierarchical remote sensing information in surface mining system”(4117134). In this paper, the manganese, lead-zinc concentration area in Huayuan County of Hunan Province as the study area. It has launched analytical work toward the preferred methods for information extraction feature of mine development status and feature distribution situation,based on medium resolution Landsat image and high resolution IKONOS image. There are the following:(1)Combined with investigation and analysis, the remote sensing interpretation keys of the mine surface elements are built. In the study area, field sampling investigations and studies are carried out to analyze and summarize the surface features of manganese and lead-zinc clusters in Huayuan County. Based on surface features out of filed survey and the priori knowledge, remote sensing image interpretation keys of the mine surface elements are built, including medium resolution Landsat image and high resolution IKONOS image. Finally, it can provide the investigation on mineral development situation and feature extraction of mine surface elements with some references.(2)This chapter analyzed the spatial-temporal changing characteristics and driving force of the Lead-zinc Mining Land in Huayuan. Based on the interpretation, five Landsat images with good quality are selected from 1990 to 2014. Combined with RS decision tree classification and GIS spatial analysis technology, adopting dynamic change model of mining regions and object transformation matrix, the chapter analyzed the land use and cover changes of mining area, evolution of spatial pattern and driving force of the lead-zinc mineralization concentrated area in Huayuan country in recent 25 years, It shows that: 1) Three mines have built since 1999 years, the land of lead-zinc mining is in a trend of sustained growth, and mainly comes from woodland. 2) The direction and scope of the lead-zinc mine expansion are bounded by ore-forming geological conditions. 3) Mining land use are subject to the double impact of the growth of socio-economic factors and government policy, the state of the three stages of change is accelerating the process to slow down its growth rate was " L " situation, in1999-2003, 2003-2009, 2009-2014. 4) Development of Mineral resources is driven by the economic benefits. In addition, the government policies, as a means of macroeconomic regulation, can balance the interests and promote mine orderly development and rational utilization in the future.(3) Taking lead-zinc mine cluster in Bamao village as a typical example, the paper aims at study how to extract the features of mine surface from high resolution remote sensing image. Taking high resolution IKONOS remote sensing images in August 2012 as based data, collecting ground object samples, analyzing features of the spectrum and texture feature of the mine surface elements, the surface elements classification rules of lead-zinc mine in Bamao village are established by adopting the object-oriented hierarchical segmentation method. According to the rules, the mine surface elements are extracted in layers. Finally, The method is compared with the traditional classification method based on pixels, the results show that object-oriented classification method has higher precision than the traditional supervised classification method, and it can well extract mine elements form high resolution remote sensing image of the typical mine region. With the help of man-machine interactive interpretation method, the classification results can be improved and the distribution images of lead-zinc mine surface elements could be got in the end(4)Based on mixed programming of ENVI/IDL 、 Arc GIS and C#, function modules are built, including the module of extraction of information area upon traditional classification method, the model of extraction of mine surface elements upon object-oriented classification method and the model of extraction upon artificial interactive interpretation method. Thus, that can give functional support for the establishment of the subsequent mine information extraction system.