| 内容提要: |
Information about the distribution and dynamics of seaport elements is critical to supporting decisions in all areas of energy supply, land use planning, risk assessment and disaster management. However, since there are various methods for estimating the current distribution of elements, further developments in the past are needed to better understand the urban process. Therefore, this paper discusses the use of common geographic data from national mapping and cadastre institutions and how to accurately reconstruct the historical distribution of oceanic elements. To this end, a method for automatically deriving such information is proposed. To this end, methods of image processing, machine learning, change detection, and partitioned gravity maps are applied. Through differential analysis, we are able to quantify errors in different types of seaport structures. |