Deciphering of emergency construction objects using satellite imagery and sub-satellite monitoring data in the Arctic

Abstract

Introduction. This paper studies depressed construction sites and their presence on the territory of Arctic. Application of earth remote sensing technologies from space is indispensable for providing sub-satellite monitoring to identify emergency, damaged and abandoned construction objects in hard-to-reach regions. The purpose of the study is the possibility of deciphering depressed construction objects according to aerospace monitoring data. For the Arctic territories remote methods are relevant because of unfavorable meteorological conditions of contact methods, as well as because of the depressed nature of most settlements. Depressed construction sites are one of the main features of the surveyed territories. In the world practice, there are certain methods for deciphering depressed structures. These are hierarchical deep learning method based on Google Street View images, information modelling of historical buildings, photogrammetry using UAVs, 3D shooting.Materials and methods. The research is carried out on the basis of satellite images of high spatial resolution, depicting territories with different lighting conditions, landscape and component composition of the Arctic surface. The subject of the research is a complex method of visual decoding of depressed construction objects.Results. The areas and signs of deciphering, the relevance of deciphering of these objects in the Arctic region are presented. Examples of emergency and abandoned objects and their deciphering signs on satellite, ground and aerial photographs are given. The ecological aspect of depressed construction objects associated with the production of landfills and certain mechanisms of behavior in relation to land use is shown.Conclusions. The methods of interpretation of depressed construction objects based on aerospace monitoring data considered in the paper allow to carry out their cadastral registration, mapping and systematization, to estimate quantitative and qualitative characteristics of these objects and depressiveness of the regions under study. This is most relevant for the Arctic region.

Keywords

• building object
• emergency construction
• destroyed object
• abandoned building
• 3-d model
• recognition
• decoding
• satellite image
• space monitoring
• semantic segmentation
• arctic