Aviation infrared video multi-target detection and tracking method and device

Abstract

The embodiment of the invention provides an aviation infrared video multi-target detection and tracking method and device. The method comprises the steps of obtaining an aviation infrared video; detecting the positions of multiple targets in the aviation infrared video by using a self-adaptive threshold method to obtain the positions of the multiple targets in the aviation infrared video; and tracking the positions of the multiple targets in the aviation infrared video by using a kernel correlation algorithm, and obtaining the motion tracks of the multiple targets in the aviation infrared video. According to the embodiment of the invention, the multi-target detection and tracking of the aviation infrared video are skillfully combined, and the positions of the multiple targets are independently stored by adopting a certain data storage structure, so that the automatic detection dynamic detection and tracking of the multiple targets in the aviation infrared video are realized.

Description

technical field [0001] Embodiments of the present invention relate to the field of computer vision, and more specifically, to a method and device for multi-target detection and tracking in aerial infrared video. Background technique [0002] The main purpose of target tracking is to calculate the two-dimensional coordinate position of the moving target on each frame of the image from the image sequence obtained by the imaging sensor through processing and analysis, and to associate the same moving target in consecutive frames in the image sequence to obtain The complete trajectory of the moving target. Simply put, it is to find the definite position of the target in the next frame of image and feed it back to the tracking system for tracking. [0003] On the one hand, in the prior art, for an infrared image, the target generally corresponds to a brighter area in the image, that is, the area of ​​interest is generally the brighter part of the image, that is, the part with hi...

AI Technical Summary

Problems solved by technology

Among them, the Kalman filter tracking algorithm is based on the target's motion information to predict and estimate the target's position to achieve tracking. It has certain robustness to the target occlusion problem, but it is only applicable to linear systems, so it is not applicable; The color probability distribution of the next frame is obtained, and the target tracking is realized through repeated iterations. The implementation process is simple and the calculation speed is fast. However, because the histogram feature is slightly deficient in the description of the target color feature and lacks spatial information, it is vulnerable to Image background; the target tracking algorithm combined with Kalman and meanshift has improved the shortcomings of these two target tracking algorithms. Therefore, it is not applicable; the particle filter tracking algorithm is generally not considered because of the huge amount of calculation

At the same time, the above method is usually used to track a single target, which cannot meet the tracking requirements of multiple targets. As a novel tracking algorithm, the tracking algorithm of kernel correlation filtering has greatly improved the tracking accuracy and speed compared with the above tracking algorithm. , and can be applied to the tracking of multiple targets in aerial infrared video

[0005] Therefore, in the prior art, there is a lack of a method and device for multi-target detection and tracking of aerial infrared video, which can combine multi-target detection of aerial infrared video with multi-target tracking of aerial infrared video to realize multi-target detection of aerial infrared video with tracking

Images