Real-time foreground extraction method based on monocular platform and device thereof

A foreground extraction and single-purpose technology, applied in the field of image processing, can solve problems such as low efficiency and instability, high hardware requirements, and long processing time, and achieve the effect of preventing background learning errors and reducing the amount of algorithm calculation

Active Publication Date: 2017-11-14
WUXI YSTEN TECH
View PDF8 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0013] Aiming at the problems of high hardware requirements, long processing time, low efficiency and instability of existing foreground extraction and recognition algorithms, a real-time foreground extraction method based on a monocular platform is proposed, including the following steps:

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
  • Real-time foreground extraction method based on monocular platform and device thereof
  • Real-time foreground extraction method based on monocular platform and device thereof
  • Real-time foreground extraction method based on monocular platform and device thereof

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0068] like figure 1 As shown, the present embodiment of the present invention provides a real-time foreground extraction method based on a monocular platform, comprising the following steps:

[0069] S110 acquires a monocular video frame sequence image;

[0070] S120 shrinks the video frame sequence image to obtain a reduced size sequence image; by obtaining frame image sequence data and reducing the image to a fixed size, the main calculation amount of the algorithm is controlled.

[0071] S130 extracting a target foreground image according to the reduced-size sequence image;

[0072] S140 performing vector edge enlargement processing on the target foreground image to obtain the foreground vector edge of the original resolution;

[0073] S150 performs anti-aliasing processing on the edge of the foreground vector in the original resolution;

[0074] S160 fills the inside of the foreground vector edge with the foreground color, fills the outside of the foreground vector edg...

Embodiment 2

[0108] Figure 5 Shows the general flow of the real-time foreground extraction algorithm based on the ARM platform according to the embodiment of the present invention, Image 6 The main algorithm modules included in the system are shown, and the specific steps are as follows:

[0109] Step 1: Image reduction, see Figure 5 , Image 6 S1 in. Scale each frame image to a fixed size by obtaining frame image sequence data. This can ensure that the amount of data is basically constant, and the calculation amount of all subsequent operations is guaranteed to be fixed.

[0110] Step 2: Background learning, see Figure 5 , Image 6 S2 in. By obtaining frame image sequence data, counting the probability p, mean value μ and variance σ of each pixel value at each position, using p, μ and σ as a model, and learning multiple models as the background statistical model of the position. A single model refers to the background model of each position in the image, where the model includ...

Embodiment 3

[0161] like Figure 9 As shown, based on the real-time foreground extraction method in the above embodiment, another aspect of the present invention also provides a real-time foreground extraction device 100 based on a monocular platform, including:

[0162] The video frame acquiring unit 110 is configured to acquire a monocular video frame sequence image;

[0163] The image reduction unit 120 is configured to perform reduction processing on the video frame sequence image to obtain a reduced scale sequence image;

[0164] The foreground extraction unit 130 is configured to extract a target foreground image according to the reduced-size sequence image;

[0165] The edge magnification unit 140 is configured to perform vector edge magnification processing on the target foreground image to obtain the foreground vector edge of the original resolution;

[0166] The anti-aliasing unit 150 is configured to perform anti-aliasing processing on the edge of the foreground vector of the ...

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 a real-time foreground extraction method based on a monocular platform and a device thereof. The method comprises the following steps of acquiring a monocular video frame sequence image; reducing the size of the image, and reducing a majority of calculation amount through reducing the size of the image; extracting a target prospect by means of a sequence frame statistics method; eliminating isolated noise by means of modes such as mean filtering and median filtering; connecting broken parts by means of a morphological method, and obtaining each target blob; extracting the contour of each blob and eliminating holes; eliminating false targets according to a blob characteristic; repairing a damaged edge by means of a specific filter; performing background updating by means of a block updating strategy; amplifying the vector edge of the target foreground, and obtaining an original resolution foreground vector edge; eliminating sawteeth which are caused by amplification; and filling the inner part of the vector edge by a foreground color, filling the rest part by a background color, and outputting the foreground image with the original resolution size. The invention further discloses a real-time foreground extraction device based on the monocular platform.

Description

technical field [0001] The present invention relates to the field of image processing, in particular to a monocular platform-based real-time foreground extraction method and a monocular platform-based real-time foreground extraction device. Background technique [0002] Both the background and the foreground are relative concepts. Take the highway as an example: sometimes we are interested in the cars coming and going on the highway. At this time, the car is the foreground, and the road surface and the surrounding environment are the background; sometimes we are only interested in the intrusion Pedestrians on the highway are interested, where the intruder is in the foreground and other things, including cars, are in the background. [0003] At present, the development of image processing technology based on PC is becoming more and more mature, but due to the large size of the PC itself, poor portability, and poor long-term running stability, the scope of application is limit...

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): G06T7/13G06T7/194G06T7/181G06T7/136
CPCG06T7/13G06T7/136G06T7/181G06T7/194G06T2207/10016G06T2207/20024
Inventor 黄飞侯立民谢建田泽康邓卉危明
Owner WUXI YSTEN TECH
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
Try Eureka
PatSnap group products