| Whether water resources in the ecological environment improvement or sustainable social development or flood disasters,are a very important part of human social life,and remote sensing technology with convenient data access,high image resolution,covering a wide range of scenes and so on,in the water information extraction has an unparalleled advantage.Therefore,the use of remote sensing technology to achieve effective monitoring of the waters,the day-to-day management of the waters,water resources planning,flood warning and flood relief and other aspects of great significance.Because the transformation of water resources is often closely related to weather changes,especially flood disasters are often accompanied by rainy weather,compared with optical satellites,synthetic aperture radar(Synthetic Aperture Radar,SAR)as an active microwave imaging sensor,the imaging process will not be affected by weather and light,in cloudy,rainy and cloudy weather can still obtain effective data,so SAR has become one of the main data sources in the field of remote sensing waters detection.In view of the existing water detection methods and the application of multi-source SAR in water detection,the following research work has been carried out in this dissertation:1.Starting from the principle of SAR imaging and imaging geometry,the water characteristics in SAR image are analyzed,and the interference factors in water detection are analyzed from the aspects of relevant spots and similar objects,and then the commonly used water detection methods are briefly summarized,and the shortcomings are pointed out.This part lays the theoretical foundation for the subsequent water detection.2.In view of the problem that accurate and fast water drawing is limited by data source,an improved single-source single-hour water detection method is proposed.In the improved method,the super-pixel segmentation algorithm is introduced first,and the edge accuracy of the detection results is effectively increased by regional growth processing,and secondly,a interference culling method based on multi-featured union is proposed by analyzing the performance of different objects in THEA image and counting the characteristic parameters of the various places.Through the water detection test of radar data of three different satellites,it is found that the improved water detection method is compared with the traditional method,on the basis of guaranteed processing efficiency,the detection accuracy has been improved to meet the needs of water detection in various unexpected situations,and to provide technical support for the subsequent multi-source SAR data water detection.3.Considering that only rely on a single data source for water detection,imaging quality is easy to bring inevitable impact on detection,this paper by analyzing the differences between on-board SAR data sources,explained the feasibility of multi-source SAR data used in water detection.In view of the limitation of single-source SAR data,a water detection method based on multi-source SAR data is proposed,the core idea of which is: to fuse the results of multi-source SAR data water detection.Finally,through two sets of multi-source data for method verification,the results show that the method can effectively integrate multi-source SAR data information,and significantly reduce shadow interference in the convergence of lift rail data.4.For the application of water detection,a dynamic monitoring method of rivers based on multi-phase SAR is designed,which is used to monitor river conditions in Xi'an and surrounding areas by using 30 Sentinel-1A satellite imagery from 2016 to 2018,and the monitoring results are compared with precipitation statistics over the years,and the monitoring results are found to be in line with precipitation trends and the reliability of this method is verified.Finally this paper uses the language of C++ to realize the engineering of water detection,and the measured data test shows that it meets the needs of engineering applications. |
PDF Full Text Request