Method and device for video multi-target tracking based on fuzzy logic

Abstract

The invention discloses a video multi-target tracking method and device based on fuzzy logic. The method includes: performing online target motion detection on the current video frame, and using the detected possible moving objects as observation results; Data association, wherein the prediction result is obtained by at least predicting the trajectory of the target in the previous video frame; trajectory management is performed on the prediction results and observation results that are not associated, including obtaining the termination trajectory segment by using the prediction results that are not associated And use the unassociated observation results to obtain new trajectory segments, and perform trajectory association on the terminated trajectory segment and the new trajectory segment. Through the above method, the present invention can effectively improve the correct association between multiple targets and observations, greatly reduce the number of target label changes in multiple target tracking, and has strong robustness and accuracy.

Description

technical field [0001] The invention relates to the field of target tracking, in particular to a fuzzy logic-based video multi-target tracking method and device. Background technique [0002] As an important research branch in the field of computer vision, video multi-target tracking technology is related to many frontier disciplines, such as image processing, pattern recognition, artificial intelligence, automatic control and computer integration. It has very important practical value in the fields of navigation, virtual reality, medical diagnosis, traffic control and monitoring. [0003] However, there are still many difficulties in developing a robust and efficient multi-target tracking algorithm for video targets in complex background environments, such as mutual occlusion of targets, the number of targets, and false observations. These situations have strong randomness and uncertainty in the actual pedestrian target tracking, which cannot be well modeled by traditional...

AI Technical Summary

Problems solved by technology

[0003] However, there are still many difficulties in developing a robust and efficient multi-target tracking algorithm for video targets in complex background environments, such as mutual occlusion of targets, number of targets, and false observations.

These situations have strong randomness and uncertainty in the actual pedestrian target tracking, and cannot be well modeled with traditional probabilistic methods

Images