2.5-dimensional map automatic registration method based on image processing technology

Abstract

The invention discloses a 2.5-dimensional map automatic registration method based on an image processing technology, and belongs to the field of image processing technologies and geographic information technologies. According to the method, the same feature points between the remote sensing image and the 2.5-dimensional map are automatically extracted through recognition and matching of image feature points by adopting a computer graphics mode, and then 2.5-dimensional registration control points meeting threshold requirements are automatically generated from the feature points through algorithm optimization, so that automatic registration of the 2.5-dimensional map is realized. According to the method, an original mode depending on manual registration is optimized into a mode that a computer automatically identifies feature points and generates control points for automatic registration. According to the method, the registration precision and the registration efficiency in the 2.5-dimensional production process are improved, and the 2.5-dimensional map production progress is greatly improved.

Description

technical field [0001] The invention relates to the fields of computer graphics and geographic information science, in particular to a fast registration method for a 2.5-dimensional map. Background technique [0002] In addition to inheriting the powerful functions of 2D electronic map in data collection, analysis and processing, 2.5D electronic map has intuitive and real spatial representation ability, which makes up for the limitation of 2D GIS graphics performance. Compared with the three-dimensional electronic map, the 2.5-dimensional electronic map has a smaller amount of data and runs faster in the background. [0003] In the current 2.5D production process, the 2.5D map needs to be registered so that it has the correct geographic coordinates. The traditional registration method is to manually select feature points for registration. This method relies on the experience of the registration personnel and the accuracy of the original base map, so both the accuracy of reg...

AI Technical Summary

Problems solved by technology

The traditional registration method is to manually select feature points for registration. This method relies on the experience of the registration personnel and the accuracy of the original base map, so both the registration accuracy and registration efficiency are low. , cannot meet mass production

Images