A fast estimation method of real-time attitude of ptz surveillance camera

Abstract

The invention relates to a PTZ surveillance camera realtime posture rapid estimation method. A scene structure panorama picture is adopted as a referential image of video matching and panorama matching is conducted on a realtime video, so that the area which the realtime video corresponds to is found in the scene structure panorama picture and a single corresponding relation between the realtime video and the scene structure panorama picture is determined. Through the single corresponding relation and calculation of a posture parameter of the scene structure panorama picture, a posture parameter of the realtime video is obtained. By means of the PTZ surveillance camera realtime posture rapid estimation method, the realtime posture of a camera can be rapidly estimated, an average processing speed can reach 15-20 fps. Moreover, an average projection error is small and realtime updating of the camera posture can be achieved effectively.

Description

technical field [0001] The invention relates to a calculation method for video monitoring, in particular to a method for quickly estimating the real-time posture of a PTZ monitoring camera. Background technique [0002] The wide application of PTZ cameras in the surveillance field has brought new challenges to the AVE surveillance system. Because the user can arbitrarily perform PTZ and zoom operations of the camera, this requires quick calibration of the camera at any attitude. In the traditional method, certain manual intervention and interactive operation are required to determine the corresponding relationship between 2D and 3D. This method is difficult to meet the requirements of real-time response in the monitoring system. [0003] The following methods can solve the problem of real-time calibration of PTZ cameras: [0004] H. S. Sawhney, A. Arpa, R. Kumar et al. proposed a new solution mode in "Video Flashlights-Real TimeRendering of Multiple Videos for Immersive Mo...

AI Technical Summary

Problems solved by technology

[0005] In actual use, this method adopts the projection method, iteratively optimizes, and has a low dependence on the color characteristics of the image, and it can still ensure good accuracy when changing the PTZ parameters of the camera at a large angle, but there are also problems that the PTZ parameters change more rapidly. Larger, the longer the iterative search time will be, which will affect the real-time performance of the algorithm

Images