Circular target real-time tracking method based on images

Abstract

The invention discloses a circular target real-time tracking method based on images, wherein the method is used to solve the problem of poor tracking stability with a circular target when moved from a far position to a near position. The method comprises the following steps: at first, a Kalman filter is constructed according to a state equation and an observation equation of the circular target, the circle center coordinate and the radius obtained in the target capturing process are taken as the initial state of the filter, prediction information of the target in the current frame is established, and the search range is reduced; then, a tracking frame is constructed according to the prediction information, and images in the tracking frame are extracted through the fast Hoff gradient method to perform circular target detection to acquire the position information of the target in the current monment; and finally, the new position is taken as the observation value of the Kalman filtering process to update filtering parameters and updata target state estimation and error covariance matrix. The above steps are performed repeatedly to realize fast and effective circular target tracking in illumination variable and background confused scenes.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a method for real-time tracking of a circular target based on image processing technology. Background technique [0002] The real-time tracking technology of moving targets based on image technology has a wide range of application values ​​in intelligent transportation systems, security protection monitoring, and military target tracking. Industrial sites, military applications, medical meteorology and many other fields often need to track moving objects with circular features for further analysis and processing. [0003] The tracking method based on motion parameter estimation uses the state space to describe the motion characteristics of the tracked system, and the Kalman filter can predict and correct the motion state of the system well, and is the most widely used method. The basic idea of ​​the Kalman filter is that through the historical measurement va...

AI Technical Summary

Problems solved by technology

[0004] In practice, when a circular target moves from far to near, as the distance between the target and the camera decreases, the imaging will gradually increase, and the target solution information will continue to change. If a global search is performed for each frame of image, it will take The time spent in a single frame image is very long, and the information of many other images will inevitably be lost, which cannot meet the real-time performance requirements

Images