Research On Information Extraction Methods Of Iron Mine Geological Environment Based On Remote Sensing Technology In North Hebei

Geological environmental problems caused by mining is increasingly grim. Experienced a traditional ground survey phase,remote sensing technology has become an effective means of rapid access to mine development, mining environmental issues and regulatory supervision of mine exploitation. In the past the mine geological environment survey generally rely on human-computer interactive interpretation, which is time-consuming, high precision but low efficiency; or rely on one computer extraction technique, time-sensitive but low extraction accuracy of some elements. The study chose the typical open-pit mining of iron ore – iron in Eastern Hebei- as the study area for its outstanding mine geological environment issues- the landscape damage. Based on the characteristics of the mine environmental elements, the study established a rapid extraction methodology. choosing three RS Image of 0.5 m resolution acquired in 2012, 2013 and 2014 as the data source, the study integrated three methods of object-oriented classification, thresholding and human-computer interactive interpretation for the information extraction of mine geological environmental elements in Eastern Hebei. The study made a quantitative analysis of the variation of the mine geological environment during year 2012 to 2014. The progress is as follows:1. Based on in-depth analysis of the spectrum, texture, geometric features, distribution and scale of the major geological environment elements in study area, the study established a rapid extraction methodology as follow: 1) Extracting open pit and dump field by using object-oriented classification method; 2) Extracting tailing ponds by using threshold segmentation method; 3) Extracting transit sites, mine construction and the recovery area by using human-computer interactive interpretation; 4) Extracting the vegetation coverage of the study area by using VFC method;2. Based on the e Cognition software and ESP Plugin, the study analysed the optimal scaling division and the rule set description which are related to object-oriented information extraction based on high-resolution remote sensing images; and construct the optimal segmentation scale and the corresponding rule sets. The practical classification accuracy verified the feasibility of the scheme. The extraction accuracy of classification results proved the feasibility of this methodology;3. Base on the results, the study made a quantitative analysis of the variation of the mine geological environment during year 2012 to 2014. The results show that the mine geological environment improved year by year.The study established this methodology based on the characteristics of environment elements, and its extraction accuracy is above 90%.