Font Size: a A A

Based On Object-oriented Extraction Of The Geological Disaster Information And Analysis Of The Characteristics Of The Geologic Hazards

Posted on:2016-12-09Degree:MasterType:Thesis
Country:ChinaCandidate:Z Y ZhouFull Text:PDF
GTID:2180330461456127Subject:Geological Engineering
Abstract/Summary:PDF Full Text Request
On May 12, 2008,A magnitude 8.0 earthquake occurred in Wenchuan, making the southwest mountainous and rivers beyond changed. In addition to the earthquake directly caused heavy casualties and property loss, the secondary geological disaster of earthquake continues to threaten people’s production and life. With the use of high resolution remote sensing image, the remote sensing technology has been widely used in geological disasters.Study of geological hazards, especially large scope, regional place, has been inseparable from the using of remote sensing image. Computer automatic extraction technology has two kinds: the extraction method based on pixels, and the extraction method based on object oriented, In dealing with a high resolution image, Use the first method, there will be "with different spectrum" and "foreign body with spectrum" problems, as well as the presence of the salt and pepper effect, so in this paper, the automatic extraction based on object-oriented technology, research on earthquake geological disaster information is extracted. Main content includes the choice of the layer, the multi-scale segmentation, extracting level building, feature selection, rule was established, finally complete the extraction. Among them, the feature selection is the key, choose multiple spectral features, shape, texture feature, after adjusting threshold, established the rules were extracted from geological disasters, and selects the membership function of fuzzy classification method, finally innovative established a set of suitable for geological disaster of earthquake information extraction method.Based on the research of the study area high resolution images and field survey, the collapse, landslide, respectively, set up indoor visual interpreting marks and characteristics, Visual interpretation of earthquake geological hazard body work, also completed the information extraction. Extracted automatically by the object-oriented technology and visual interpretation methods of comparative analysis, the interpretation of the collapse area overlap degree of the landslide mass is as high as 86.27%, but there are also different, comparison and analysis the following conclusions: Time of using automatic extraction method is shorten, greatly improving the working efficiency; Automatic extraction method is more objective, reducing artificial subjective factors; In results of automatic extraction method, the shape of collapse and slide boundary looks like saw tooth, and the results of boundary of visual interpretation method is more accepted by people; the slide in sunny slope is more accurate using automatic extraction method of the study area, but the slope collapse body extraction effect is poorer, there is no such problem in visual interpretation. In practice, it is recommended that the method of automatic extraction adopted first, then modifying visual interpretation based on expert knowledge, combined with their respective advantages.According to the large number of data of geological disasters through extraction,on the study of characteristics, the first study is its types, including collapse, landslide, debris flow, In key area, collapse can be divided into three types,including shatter type, denudation type, erosion type, landslide is divided into two categories, soil and rock landslide, Followed by the study of the scale of the geological disasters, by means of remote sensing interpretation and investigation into key areas, in key areas, a total of 2067 collapse were happened after the earthquake, with 779 small collapse, 1015 medium-sized collapse, 261 large collapse, 12 oversize collapse;717 landslides, of which 535 small landslides, 162 medium-sized landslide, 20 large landslides. Finally, the collapse and landslide distribution and controlling factors were analyzed, combined collapse and landslide with the five main factors that influence and control of its distribution, Including fault, elevation, gradient, slope direction and drainage, Through the analysis,respectively,there are several characteristics in its distribution: They are mainly distributed in the vibration fault within 8 km, 1500 m to 2500 m elevation distribution area, the largest gradient between 40 ° to 80 °, area and density is the largest slope upward SE direction, most within 400 m from the drainage; there is a certain difference in collapse and landslide distribution, especially on the slope. Collapse and landslides along the highway will poses a direct threat to pedestrians and highway safety, located within the debris flow gully collapse and landslide provide sufficient source for outbreak of debris flow.
Keywords/Search Tags:Object oriented, information extraction, geological disasters, spatial distribution, Characteristics analysis
PDF Full Text Request
Related items