Video multi-target fuzzy data correlation method and device

Abstract

The invention discloses a video multi-target fuzzy data correlation method and device. The method comprises the steps of carrying out online target motion detection on a current video frame, and obtaining a possible moving object as an observation result; calculating the shielding degree among the prediction results of different targets in the current video frame; according to the shielding degree, respectively judging whether the shielding is generated between each prediction result and other prediction results or not; if no shielding is generated between the prediction result and any other prediction result, performing the first data correlation on the prediction result and the observation result; if the shielding is generated between the prediction result and any other prediction result, performing the second data correlation on the prediction result and the observation result, wherein the first data association is different from the first data association. By means of the above mode, the accuracy of multi-target tracking in a complex environment can be improved.

Description

technical field [0001] The invention relates to the field of target tracking, in particular to a video multi-target fuzzy data association method and device. Background technique [0002] Video multi-target tracking has always been a hot research issue in the field of computer vision. It integrates image processing, pattern recognition, artificial intelligence, automatic control and computer. It has very important practical value in fields such as medical diagnosis, traffic control and monitoring. [0003] However, for video targets in complex background environments, there are still many difficulties in developing a robust and efficient multi-target tracking algorithm, such as mutual occlusion of targets, sudden changes in illumination and posture, scale changes, and rapid movement of background clusters. These situations have strong randomness and uncertainty in the actual pedestrian target tracking, and cannot be modeled well with traditional probabilistic methods. Con...

AI Technical Summary

Problems solved by technology

[0003] However, for video targets in complex background environments, there are still many difficulties in developing a robust and efficient multi-target tracking algorithm, such as mutual occlusion of targets, sudden changes in illumination and posture, scale changes, and rapid movement of background clusters, etc.

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

Images