Image foreground and background segmenting method and device

Abstract

The invention provides an image foreground and background segmenting method and device, wherein the method comprises the following steps of: segmenting a target frame image into a plurality of blocks; determining the motion states of all the blocks, wherein the motion states include a static state and a dynamic state; counting the static state time of the block in the static state; refreshing a background image of the block in the static state when the static state time of the block is larger than a first preset threshold; and extracting a foreground image from the target frame image according to the refreshed background image. By using the image foreground and background segmenting method and device, the segmentation processing course of a foreground and a background of the image can be simplified, the real-time processing requirement of an embedded device is met, the image segmenting accuracy is improved, resources are saved, and the processing efficiency is increased.

Description

technical field [0001] The present invention relates to the technical field of image segmentation, in particular to a method for image foreground and background segmentation, and an image foreground and background segmentation device. Background technique [0002] With the continuous development of computer science and technology, image processing and analysis has gradually formed its own scientific system, and new processing methods emerge in endlessly. Although its development history is not long, it has attracted widespread attention from all walks of life. In the process of image research and application, people are often interested in certain parts of the image. These parts are called target or foreground images. They generally correspond to specific areas and analysis targets with unique properties in the image. These areas need to be separated. Or extracted, on this basis it is possible to further use the target. Image segmentation is the technology and process of di...

AI Technical Summary

Problems solved by technology

Doing target detection in the background model with smear will seriously affect the result of foreground segmentation

[0004] In order to solve the problem of smear generated by the Gaussian mixture model, a variety of solutions algorithms have been proposed in the prior art. Although these algorithms can effectively solve the smear problem generated when the Gaussian mixture model foreground image is segmented in most cases, when the target frame image When a target enters the monitoring range and stays still for a long time, or when there is an occasional moving target in the target frame image, the target will not be brushed into the background in time; or, when only part of the foreground image can be recognized during the foreground image segmentation process Smearing may still occur when the target or its outline

In addition, most of the algorithms in existing solutions are computationally complex and cannot meet the real-time processing needs of embedded devices

Images