An optimization system and method for a motion model

Abstract

The invention discloses an optimization system and a method of a motion model. The method comprises the following steps of: acquiring N frame image information in a camera; acquiring a state vector ofa tracked target center in a pixel coordinate system in a k-th frame image, wherein k satisfies 1 <= k <= N; The second-order autoregressive motion model is used to predict the state vector of the target center in the (k+1) th frame image in the pixel coordinate system. The jerkiness constraint conditions of the target center in two consecutive adjacent images are established. The state vector ofthe tracked target center in the (k+1) th frame image is constrained by the jerky degree constraint condition in the pixel coordinate system. Obtaining the predicted position of the tracked target center according to the predicted state vector of the tracked target center in the coordinate system; The predicted position of the tracked target center is used to modify the target position of the searching area, and the accurate center position is obtained. The invention can accurately predict the position of the tracked target center while reducing the range of the searching area, and improve the efficiency of the target tracking.

Description

technical field [0001] The invention relates to the technical field of model optimization, in particular to a motion model optimization system and method. Background technique [0002] Moving target detection and tracking is a branch of image processing and computer vision, which has great significance in theory and practice. The research of moving target detection and tracking technology is just to extract the moving target from the image data and keep tracking it. Provide basic elements and analysis basis for further video sequence images. Object detection refers to extracting the foreground object that has relative motion with the background, and object tracking refers to determining the corresponding position of the same object in different frames of the video sequence. The detection and tracking of moving targets is a key technology and hot issue in the research fields of computer vision, image processing and active surveillance. [0003] At present, in the process o...

AI Technical Summary

Problems solved by technology

[0003] At present, in the process of using the KCF algorithm to track a moving target, the fast and violent movement of the tracked target will cause the search area of ​​the original KCF algorithm to not completely cover the tracked target, which will cause tracking drift or even tracking failure.

[0004] In the existing technology, the search area is guaranteed to cover the target by expanding the area of ​​the search area, but this will increase the amount of calculation and reduce the tracking efficiency

Images