Unlock instant, AI-driven research and patent intelligence for your innovation.

A Superpixel-Based Object Tracking Method in Complex Scenes

A complex scene and target tracking technology, applied in the field of target tracking in complex scenes based on superpixels, can solve problems such as easy to lose track, difficult to deal with lighting changes, and inability to separate the target from the background

Inactive Publication Date: 2017-10-31
UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, since the local binary model cannot completely separate the target from the background, it is easy to lose track when the target is moving fast or the target is blurred.
At the same time, the standard local binary model is difficult to deal with lighting changes

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
  • A Superpixel-Based Object Tracking Method in Complex Scenes
  • A Superpixel-Based Object Tracking Method in Complex Scenes
  • A Superpixel-Based Object Tracking Method in Complex Scenes

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0066] A method for tracking objects in complex scenes based on superpixels, comprising the following steps:

[0067] Step 1: Initialize the parameters, take the first m frames of images in the video stream as the training set, and m is 3-6:

[0068] Step 1.1: Select the area where the target is located, and perform superpixel segmentation on the area:

[0069] Step 1.1.1: Randomly select k points in the target area as the center point of the superpixel, also called the cluster center C of the superpixel K :

[0070] C K =[l k ,a k ,b k ,x k ,y k ] T

[0071] Sample pixels in each regularly sized box S, l k 、a k , b k is a parameter in the CIELAB color space, which is the closest to human vision, and its l k The component is the brightness. can modify a k (red-green value) and b k (Blue-yellow value) component output color scale to do accurate color balance, which is more accurate than RGB space. x k 、y k Represents the coordinates of a point. S can be take...

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 discloses a target tracking method in a complex scene based on superpixels, which includes the following steps: (1) initializing parameters, taking the previous frame image in the video stream as a training set, taking 3-6; (2) finding the above For the target area of ​​a frame, repeat the above steps, use superpixels for segmentation, and extract superpixel features; (3) calculate the confidence map of superpixels; (4) use the confidence map obtained from the previous frame to take an area as a candidate for the target area; (5) Based on Bayesian theory, the probability of the area where the target is located in the current frame can be obtained; (6) Update the training set and appearance model; A variety of complex environments, such as illumination changes, target pose changes, scale changes, target occlusion, etc., and when there are blurred frames in the video sequence, it can still track accurately.

Description

technical field [0001] The invention relates to a superpixel-based target tracking method in complex scenes. Background technique [0002] Object tracking technology in video is a key technology of video image processing. It involves the fields of computer vision, pattern recognition and artificial intelligence, and has important theoretical significance. Most of the scenes in real life will change, and many existing target tracking algorithms cannot meet the target tracking in real complex scenes. Target tracking based on complex scenes needs to comprehensively consider a series of complex changes in the video sequence, the surrounding environment of the target to be tracked, and the target itself, including target blocking, target loss, illumination changes, rapid target movement, target attitude changes, and target scale. Variations and other complex factors. [0003] Tracking failure is a major problem faced by object tracking in complex scenes. For example, tracking...

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 Patents(China)
IPC IPC(8): G06K9/00G06K9/62
CPCG06V20/46G06V20/49G06V2201/07G06F18/253
Inventor 蒲晓蓉陈雷霆张思远邱航蔡洪斌崔金钟卢光辉曹跃
Owner UNIV OF ELECTRONICS SCI & TECH OF CHINA