Unmanned aerial vehicle-based large scene multi-modal analysis system and method

Abstract

The application relates to the technical field of aerial video analysis, and discloses a large-scene multi-modal analysis system and method based on a UAV. The method acquires an original video frame sequence from aerial equipment, extracts a scene change period and an object motion trend by calculating an inter-frame motion feature vector, and judges whether a scene is in a stable or dynamic state by means of a state classification model. In the dynamic state, a tracking target position is calculated according to preset parameters, and an analysis target area is adjusted in combination with the motion trend; a camera parameter adjustment instruction is generated according to a deviation amount of the analysis target area from an actual object position, and is output to an executing mechanism; a complexity value is acquired through a scene complexity feedback signal, and when the complexity value exceeds a preset range, an adaptive algorithm is used to update an analysis strategy and output an adjustment signal; the output timing of control and strategy adjustment signals is optimized, the frame sequence and analysis parameter change trend are monitored in real time, preset parameters are dynamically updated to match user requirements, and efficient video analysis processing is realized.

Images