An automatic reconnaissance system for unmanned aerial vehicle photoelectric stabilization platform based on image matching

Abstract

The invention discloses an image-matching-based automatic reconnaissance system of an unmanned aerial vehicle mounted photoelectric stabilization platform. According to the system, route planning and automatic scanning control for automatic scanning are implemented according to the image registration relationship of photoelectric load in large and small view fields, so that automatic traverse scanning on images of a preset observation area is completed; in automatic scanning, images of adjacent image sequences can be matched, motion compensation can be implemented according to the matching relationship, and frame difference calculation can be also implemented, so that a moving target can be detected, real-time automatic recognition on the moving target in an image can be achieved, and the reconnaissance accuracy and efficiency are improved.

Description

technical field [0001] The invention relates to an automatic reconnaissance system for an unmanned aerial vehicle-mounted photoelectric stabilization platform based on image matching, and belongs to the technical field of unmanned aerial vehicle reconnaissance. Background technique [0002] While the UAV-borne optoelectronic stabilization platform provides mechanical stabilization functions for optoelectronic payloads, it must achieve long-term precise observation of the ground observation area under a small field of view. In order to quickly find suspicious moving targets, observers also need to identify moving targets in images. [0003] In order to achieve high-resolution viewing of the preset viewing area, currently the optical axis of the platform is manually locked to the viewing area under a large field of view, and then the camera equipment is adjusted to a small field of view, and the angle of the optical axis is manually increased to change the angle. Take aim. T...

AI Technical Summary

Problems solved by technology

The current solution is generally to manually adjust or use the control unit to add scanning function to the photoelectric stabilized platform, but the scanning is only relative to the platform's own coordinate system, and it cannot judge whether the image captured by the platform's current attitude exceeds the observation area. The scan feature for is not available

[0004] The main purpose of the UAV-borne photoelectric stabilization platform is to accurately observe the ground observation area, and to quickly find and identify suspicious moving targets. However, it is difficult for the human eye to respond quickly to moving targets in the image when the video shakes, and Visual fatigue caused by jitter greatly reduces the efficiency of scouting

Images