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

A Foreground Segmentation Method Based on Binocular Stereo Vision System

A technology for binocular stereo vision and foreground segmentation, which is applied in image analysis, instrumentation, computing, etc. to achieve a wide range of applications and improve accuracy.

Active Publication Date: 2020-07-28
UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF7 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, several active contour models that have been proposed are limited to segmenting monocular grayscale images and cannot be well applied to binocular color images.

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 Foreground Segmentation Method Based on Binocular Stereo Vision System
  • A Foreground Segmentation Method Based on Binocular Stereo Vision System
  • A Foreground Segmentation Method Based on Binocular Stereo Vision System

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0082] In this embodiment, a foreground segmentation method based on a binocular stereo vision system uses a binocular stereo vision system to obtain a left view and a right view of the same object, in order to distinguish the foreground of the left view from the background and separate the foreground of the right view To distinguish from the background, an energy equation is defined, and the energy equation is used to assign different labels to the foreground pixels in the foreground and the background pixels in the background, so as to distinguish the foreground from the background. This scheme assigns different labels to foreground pixels and background pixels in the picture by defining an energy equation, so that foreground pixels and background pixels can be accurately distinguished, improving the accuracy of foreground segmentation, and this scheme can achieve The foreground segmentation of binocular color images makes this scheme widely applicable. And the process of fo...

Embodiment 2

[0084] On the basis of the foregoing embodiments, in the present embodiment, the energy equation is:

[0085] E=∑ i P(c i , l i )+∑ i，j S(c i , l i ; c j , l j ) (1)

[0086] where: c i and c j are adjacent pixels,

[0087] l i is the pixel point c i The corresponding binary label,

[0088] l j is the pixel point c j The corresponding binary label,

[0089] P is the data item,

[0090] S is the smoothing term,

[0091] E is the energy value,

[0092] The optimal segmentation result can be obtained by making E the minimum value.

[0093] By defining the energy equation to assign different labels to the foreground pixels and background pixels in the picture to obtain the segmentation result, so that the obtained segmentation result satisfies the data item and the smoothing item at the same time, making the segmentation result correct, and can effectively avoid the existing technology. Holes and discontinuities that are easy to appear in other segmentation met...

Embodiment 3

[0095] On the basis of the above-mentioned embodiments, in this embodiment, optimization processing is also included on the foreground pixels and the background pixels, and the optimization processing includes the following steps:

[0096] Step S1: Calculate the original disparity map. According to the original disparity map of the left view and the right view, since the disparity value of the foreground is greater than the disparity value of the background, the foreground pixels and background pixels in the original disparity map can be distinguished according to the original disparity map , but the original disparity map has noise, holes, and even wrong disparity values ​​at the edge of the object, such as figure 1 As shown in (c), the difference between foreground pixels and background pixels is not very obvious.

[0097] Step S2: In order to remove noise points, fill hole points, adjust wrong disparity values, and improve the clarity and integrity of the original disparit...

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 foreground segmentation method based on a binocular stereo vision system. The binocular stereo vision system is used to obtain the left view and the right view of the same object, distinguish the foreground of the left view from the background, and separate the foreground and background of the right view. To distinguish, define an energy equation, and use the energy equation to assign different labels to the foreground pixels in the foreground and the background pixels in the background, so as to distinguish the foreground from the background. The beneficial effects of the present invention are: the present invention assigns different labels to the foreground pixels and the background pixels in the picture by defining an energy equation, so that the foreground pixels and the background pixels can be accurately distinguished, and the accuracy of the foreground segmentation is improved. degree, and this scheme can realize the foreground segmentation of binocular color images, which makes this scheme widely applicable.

Description

technical field [0001] The invention relates to the fields of computer vision and computer graphics, in particular, it is a foreground segmentation method based on a binocular stereo vision system. Background technique [0002] With the advancement of science and technology, binocular stereoscopic images gradually occupy a place in people's lives, and their applications in various fields are becoming more and more important. Such as object tracking, automatic navigation, medical aided diagnosis, virtual reality, map drawing, etc. Image engineering can usually be divided into three levels, image processing, image analysis and image understanding. As a key step in the process from image processing to image analysis, image segmentation has been the focus and difficulty of research for a long time. In recent years, the active contour model has become a hot spot in the field of segmentation because of its advantages of easy modeling and efficient mathematical solution. This ty...

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): G06T7/194G06T7/90
Inventor 刘帅成杨涛涛孙超曾兵
Owner UNIV OF ELECTRONICS SCI & TECH OF CHINA