Research On Application Of UAV-borne Remote Sensing In The Investigation And Analysis Of Ancient Village

Cultural heritage is the crystallization of ancient people,which reflects the development of human society.It is of high importance to being protected,as its prominent functions in cultural inheritance and development.But urbanization is eroding and reshape the traditional context with an ever-increased rate.Heritage landscapes are being pushed to the brink of extinction because of natural and anthropic aspects.Meanwhile,stereotypical methods are time-consuming,unsafe and unstable,which would affect the data integrity,stability and accuracy.Therefore,searching a reliable,efficient and easy-operated data acquisition and analysis technique is on the agenda for the territory of heritage landscape protection.According to the needs of heritage protection,we present the application of low-altitude UAV remote sensing technique for data acquisition in this paper.Based on the image data,relevant research is conducted on the strategy for ancient landscape recognition and extraction.Furthermore,the application of the analysis result in the heritage protection is also discussed.In the paper,we take traditional landscapes of an original village as research object.The village is located in mountainous area.Based on the characteristics of landscape,a categorical regulation satisfying the needs of heritage protection is presented,also,a feature space for the landscape extraction is established.Moreover,research on heritage landscape extraction is conducted by utilizing the combination of Analytic Hierarchy Process(AHP)and object-oriented image analysis technique(OBIA).In addition,the output is evaluated by using the results of situ measurement,and its applications are further discussed.The main study content can be listed as follows:(1)a top-down,concise and high-efficiency operated categorical regulation is established and applied to in the case study.(2)In the light of the process for remote-sensed data analysis,the Hierarchical model for heritage landscapes is established by using AHP,OBIA,image segmentation and classifiers(SVM).The trained model is assessed by the ground truth,and its accuracy reaches 70%at least.In addition,method comparison is conducted on the image segmentation and classification.The result shows that method used in the paper is more stable and accurate.(3)Based on the extraction result,spatial analysis of traditional landscapes is carried out,which mainly includes area and distance statistics and Landscapes status is also analyzed and discussed.The successful application the UAV remote sensing technique presented in the paper satisfies the need of digital heritage protection.It has advantages including high efficiency,strong reliability and less limitation,compared with traditional methods.Such method has practical application value,which can simplify and improve heritage protection work to large extent.