Long-time sheltering robust tracking method based on convolutional features and global search detection

A global search and convolution technology, applied in the field of computer vision, which can solve problems such as tracking failure and appearance model drift.

CN106952288AActive Publication Date: 2017-07-14NORTHWESTERN POLYTECHNICAL UNIV
3 Cites 22 Cited by

Patent Information

Authority / Receiving Office
CN · China
Current Assignee / Owner
Publication Date
2017-07-14

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

The invention relates to a long-time sheltering robust tracking method based on convolutional features and global search detection. Through adopting a convolutional feature and multi-scale relative filtering method in a tracking module, the feature representation capacity of a tracking object appearance model is enhanced, so that a tracking result is very robust to factors such as illumination variation, object scale variation and object rotation. In addition, through the introduced global search detection mechanism, a detection module can detect the object again when the tracking fails because the object is sheltered for a long time, thus the tracking module can recover from error, and the object can be continuously tracked for a long time even in the case of object appearance changes.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The invention belongs to the field of computer vision and relates to a target tracking method, in particular to a long-term occlusion robust tracking method based on convolution features and global search detection. Background technique

[0002] The main task of target tracking is to obtain the position and motion information of a specific target in a video sequence, which has a wide range of applications in video surveillance, human-computer interaction and other fields. During the tracking process, factors such as illumination changes, complex background, target rotation or scaling will increase the complexity of the target tracking problem, especially when the target is blocked for a long time, it is more likely to cause tracking failure.

[0003] The tracking method (TLD for short) proposed in the document "Tracking-learning-detection, IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012, 34(7): 1409-1422" combines the traditional t...

Examples

Embodiment Construction

[0043] Now in conjunction with embodiment, accompanying drawing, the present invention will be further described:

[0044] Step 1 Read the first frame of image data in the video and the initial position information of the target [x, y, w, h], where x, y represent the abscissa and ordinate of the target center, w, h represent the width and high. The coordinate point corresponding to (x, y) is marked as P, and the target initial area with a size of w×h is marked as R with P as the center init , and then record the scale of the target as scale, initialized to 1.

[0045] Step 2. Taking P as the center, determine a region R containing target and background information bkg , R bkg The size of is M×N, M=2w, N=2h. Using VGGNet-19 as the CNN model, the convolutional feature map z is extracted from R' in the 5th convolutional layer (conv5-4) target_init . then according to z target_init Build the target model for the tracking module t∈{1,2,...,T}, T is the number of CNN model ...