Video tracking method based on multi-stage characteristics

Abstract

The invention relates to a video tracking method based on multi-stage characteristics, which comprises the following steps of: 1, selecting a target to be tracked in a frame image, and extracting characteristics of the target as a template characteristic; 2, finding out an estimation target in the current frame image according to the template characteristic; 3, judging the estimation target, and controlling the updating of the multi-stage template characteristic by using a multi-stage updating threshold; 4, determining the target to be tracked; and 5, circularly executing the steps from 2 to 4 until all frame images are completely processed. The invention has reasonable design, and is used for controlling updating the multi-stage template characteristic by using the multi-stage updating threshold; when a certain stage template characteristic is judged that the target is lost, the next stage template characteristic is used for updating the template characteristic, thereby ensuring that the target is effectively tracked under the conditions that the environment is changed seriously and the target is shaded, greatly enhancing the effectiveness of tracking the target and enlarging the application range.

Description

technical field [0001] The invention belongs to the field of video image processing, in particular to a video tracking method based on multi-level features. Background technique [0002] Visual target tracking processing is an important part of computer vision processing, and visual target tracking processing is also an important research direction in digital video technology, from industrial detection to security monitoring, from medical images to military and many other fields. It has a wide range of applications, such as the detection of traffic flow, the security of important places, the automatic driving or assisted driving of aviation and automobiles, and the aspects of weapon guidance and control in the military. Due to the diversity of tracking target forms and the uncertainty of target feature changes, how to achieve effective target tracking in various environments has always been a difficult problem. The existing video target tracking algorithm usually adopts a s...

AI Technical Summary

Problems solved by technology

The existing video target tracking algorithm usually adopts a single update threshold to control the template feature update, as shown in Figure 3, the existing problem is: when the target environment changes drastically and the target is occluded, due to the target The feature change is relatively fast, and the update of the target template feature cannot keep up with the speed of the target feature change, resulting in the phenomenon of target loss or tracking error, which greatly limits the effectiveness of tracking

Images