Target tracking method and device in video monitoring

Abstract

The invention provides a target tracking method and a target tracking device in video monitoring. The method comprises the steps that the state value corresponding to a target in a previous frame is obtained; a current frame is subjected to target detection, and the observed value corresponding to the target observing in the current frame is obtained; a cost matrix is built according to the state value and the observing value, in addition, an EMD (earth mover's distance) algorithm is applied for solving the cost matrix, and an allocation matrix is obtained; and the target in the current frame is recognized according to the allocation matrix. The device comprises an obtaining module, a target detecting module, an operation module and a recognition module, wherein the obtaining module is used for obtaining the state value corresponding to the target in the previous frame, the target detecting module is used for carrying out the target detection on the current frame to obtain the observed value corresponding to the target observing in the current frame, the operation module is used for building the cost matrix according to the state value and the observing valve and applying the EMD algorithm for solving the cost matrix to obtain the allocation matrix, and the recognition module is used for recognizing the target in the current frame according to the allocation matrix. When the target tracking method and the target tracking device are adopted, the accuracy can be improved.

Description

technical field [0001] The invention relates to intelligent video monitoring technology, in particular to a method and device for tracking objects in video monitoring. Background technique [0002] In intelligent video surveillance, multi-target tracking will be a key link, and multi-target tracking for dense targets will become a challenging problem. The core of multi-target tracking lies in the data association problem. Traditional methods for multi-target tracking include local nearest neighbor association, global nearest neighbor association, multi-hypothesis tracking and joint probability filtering. Most of these methods are based on radar It is borrowed from the field of target tracking and the field of sonar target tracking. [0003] However, since video target tracking is quite different from radar target tracking and sonar target tracking, it will be difficult to apply these existing methods in the field of radar target tracking or sonar target tracking to the real...

AI Technical Summary

Problems solved by technology

[0003] However, since video target tracking is quite different from radar target tracking and sonar target tracking, it will be difficult to apply these existing methods in the field of radar target tracking or sonar target tracking to the realization of target tracking in viewpoint monitoring. With large defects, the accuracy of multi-target tracking is very low

Images