Multi-geographic-video adaptive event detection method based on abnormal change modeling

The invention relates to a geographic video adaptive event detection method based on abnormal change modeling, which belongs to the technical field of geographic space data processing, and comprises the following steps: a) constructing a cognition-driven geographic video abnormal change model; b) constructing a geographic video anomaly hierarchical perception structure; c) creating a geographic video abnormal event measurement index; and d) calculating a geographic video abnormal event measurement index. The core of the invention is to establish a self-adaptive event detection method of a geographic video aiming at organization and retrieval requirements of rapid integration and efficient knowledge understanding of the geographic video reflecting multi-scale complex public security event information, so as to realize perception and detection of abnormal events from process segments of discrete multi-geographic video contents. 本发明涉及一种基于异常变化建模的地理视频自适应事件探测方法，属于地理空间数据处理技术领域，包括以下步骤：a)构建认知驱动的地理视频异变模型；b)构建地理视频异常分层感知