In order to quickly extract and distribute the incremental information of spatial database, this work analyses the reasons of incremental information of spatial database with the instance of residential feature. Firstly, a new concept of increment is proposed based on the spatio-temporal change type and data delta. The change type describes the semantics of increment information and data delta describes the content of increment information. Based on this concept, the types of residential change and relevant data deltas are put forward. There are two catalogs of change type, i.e. subjective change and objective change. The subjective change can be detailedly divided into 3 small change types and the objective change can be detailedly divided into 12 small change types. Additionally, the incremental information resulted from the objective change is formalized. Based on the above formalized changes, the automatic extraction of incremental information can be expediently programmed. The classifying result was applied to the incremental information automatic extracting system of 1.250,000 topographic databases.