Video target tracking method, system and device and storage medium

Abstract

The invention discloses a video target tracking method, system and device and a storage medium. The method comprises the following steps: intercepting a picture frame from a video shot by a camera, using a mean shift algorithm and a Kalman filtering algorithm to carry out target identification on the image frame; carrying out weighted average on an identification result of the mean shift algorithm and an identification result of the Kalman filtering algorithm; and determining the position of the target in the picture frame according to the result of the weighted average, and controlling the holder to drive the camera to move, thereby enabling the target to be located in the central area of the picture frame. The position of the target in the visual field of the camera is determined through the identification result of the mean shift algorithm and the weighted average result of the identification result of the Kalman filtering algorithm, the identification precision is high, and the method, system and device can adapt to different backgrounds and is suitable for video target tracking of a non-fixed camera. The method, system and deviceare widely applied to the technical field of image processing.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a video object tracking method, system, device and storage medium. Background technique [0002] In the fields of human-computer interaction and video surveillance, it is necessary to lock and track specific targets such as faces or license plates in videos. The existing video target tracking technology is mainly applied to fixed cameras, and when it is applied to non-fixed cameras driven by equipment such as pan / tilt, since the field of view of the non-fixed camera may change, the existing video target tracking technology may The inability to adapt to the frequently changing field of view background results in the inability to accurately track video targets. Contents of the invention [0003] In order to solve the above technical problems, the object of the present invention is to provide a video object tracking method, system, device and storage medium. [0004] On...

AI Technical Summary

Problems solved by technology

The existing video target tracking technology is mainly applied to fixed cameras, and when it is applied to non-fixed cameras driven by equipment such as pan / tilt, since the field of view of the non-fixed camera may change, the existing video target tracking technology may Unable to adapt to the frequently changing field of view background, resulting in inability to accurately track video targets

Images