Eureka AIR delivers breakthrough ideas for toughest innovation challenges, trusted by R&D personnel around the world.

Adaptive Mean Shift target tracking method based on LBP features

A target tracking and adaptive technology, applied in the field of computer vision and target tracking, can solve the problems that color features are sensitive to illumination changes, lack of target model update strategy, and cannot effectively identify tracking targets, etc.

Inactive Publication Date: 2017-04-19
TIANJIN UNIV
View PDF2 Cites 17 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] (1) The color histogram is used as the apparent feature of the target, which cannot contain all the information of the target; and the color feature is sensitive to illumination changes, and it is easy to lose the tracking target under the condition of illumination change; in addition, when the target and the background color are similar In this case, the tracking target cannot be effectively identified (see literature [3] and [4])
[0006] (3) The scale and direction of the target cannot be effectively estimated (see literature [5])
[0007] (2) Lack of an effective target model update strategy (see literature [6] and [7])
[0008] (4) It cannot overcome the defect of background clutter interference. When there are many clutters in the background, it is easy to lose the tracking target (see literature [8])

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Adaptive Mean Shift target tracking method based on LBP features
  • Adaptive Mean Shift target tracking method based on LBP features
  • Adaptive Mean Shift target tracking method based on LBP features

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0085] A scale-direction adaptive Mean Shift target tracking method that combines local binary features and color histograms. The corresponding schematic diagram of the tracking framework is as follows figure 1 As shown, the tracking method can be divided into three steps: object model generation, object scale orientation estimation and similarity measurement.

[0086] (1) Target model generation:

[0087] In the tracking method provided by the present invention, the target model is described by a joint histogram composed of local binary features and color features of the image, that is, using the color and texture features in the mask formed by the local binary mode to describe the target, constructing A target model for joint texture-color features.

[0088] (2) Similarity measure:

[0089] The Bhattacharyya coefficient is used to measure the similarity between the target model and the target candidate model. The Bhattacharyya coefficient represents the cosine value of th...

Embodiment 2

[0094] Combine below figure 1 1. The design principle introduces the scheme in embodiment 1 in detail, see the following description for details:

[0095] A scale-direction adaptive Mean Shift target tracking method that combines local binary features and color histograms. The schematic diagram corresponding to the tracking frame is as follows figure 1 As shown, object model generation, object scale orientation estimation and similarity measure. The specific implementation manners of these three parts will be described in detail below.

[0096] (1) Target model generation part:

[0097] The description of the target model in the tracking method provided by the present invention consists of a joint histogram composed of local binary features (LBP) and color features. The following is a detailed introduction on how to extract the local binary features of the target area.

[0098] Local Binary Pattern (LBP) is an operator used to describe the local texture features of the ima...

Embodiment 3

[0157] Below in conjunction with specific accompanying drawing, the scheme in embodiment 1 and 2 is carried out feasibility verification, see the following description for details:

[0158] The tracking method provided by the implementation of the present invention is used to track the video under the similar situation of 3 groups of illumination changes and background and target color, and compare it with the tracking result of the SOAMST algorithm under the same conditions, and the obtained partial tracking results are shown in Fig. ), (3), (4) shown.

[0159] In the video sequence shown in Figure (2), the wild goose and its surrounding areas are selected as the tracking target. The recognition of the target in the video is very low, and the scale direction of the wild goose is constantly changing during the tracking process. figure 2 (a) and (b) represent part of the tracking results obtained by the tracking algorithm provided by the present invention and the SOAMST algori...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention relates to the technical fields of computer vision and target tracking, and aims to solve the problem on how to ensure the robustness of target tracking in different background interference situations, improve the robustness and adaptability of an algorithm and effectively overcome constant change of the dimension and direction of a target in the process of tracking. According to the technical scheme employed in the invention, an adaptive Mean Shift target tracking method based on LBP features comprises the following steps: (1) generation of a target model; (2) similarity measurement: the similarity between the target model and a target candidate model is measured with a Bhattacharyya coefficient; and (3) target dimension and direction estimation: Mean shift iteration is performed on a target area to make the target area converge to the spatial location of a candidate target, matrix decomposition is performed on a target candidate area weight map integrating texture and color features, and the dimension and direction of the target candidate area are calculated through matrix analysis. The method is mainly used in target tracking occasions.

Description

technical field [0001] The invention relates to the technical fields of computer vision and target tracking, in particular to a scale-direction adaptive Mean Shift target tracking method combining two-dimensional local binary features and color histograms. Background technique [0002] Computer vision is an emerging discipline developed in recent years. Its research covers intelligent monitoring systems, robot visual navigation, human-computer interaction, three-dimensional reconstruction of objects, automatic driving and other fields. Among the many research fields of computer vision, moving target tracking based on image sequences has attracted extensive attention from academic and industrial circles at home and abroad, and has important applications in the fields of intelligent monitoring, robot navigation, intelligent transportation, video content analysis and understanding, etc. Value is an indispensable key technology. [0003] A large number of excellent tracking alg...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06T7/246G06T7/73G06T7/77G06K9/62
CPCG06T2207/10016G06T2207/30221G06F18/22
Inventor 唐晨程佳佳苏永钢李碧原
Owner TIANJIN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Eureka Blog
Learn More
PatSnap group products