Urban large-scale digital mapping demanded a greatly efficient and speedy surveying method. The traditional surveying process lagged behind because it had many shortcomings in mapping cost, mapping period, data storage and further usage. It is an effective way to use high spatial resolution remote sensing image to make large-scale urban maps. Firstly this paper discussed the relationship between mapping scale and image spatial resolution, analyzed the way of processing image about large-scale mapping, explained the data-abstracted method based on customized rules, intelligent symbols and automated topology. Secondly this paper expatiated on the core technology of geo-database based on large-scale digital maps and high spatial resolution image, argued how to establish and maintain a spatial database and how to serve the other different applications by the public interface. The remote sensing image, large-scale digital maps, intelligent symbols, map elements, map-framing indexes and social statistics data were stored into the same spatial database which further provided comprehensive data and the thematic maps for urban administration, urban plan, land management, digital city and Urban GIS. And then the spatio-temporal geo-database was established by the "versioning" which could manage different spatial and temporal data. So it was feasible to produce large-scale digital maps using high spatial resolution remote sensing image and construct spatio-temporal database using "versioning." Furthermore it is a necessary way to urban informational and intelligent administration.