Moving object detection method based on multi-threshold self-optimization background modeling

Abstract

The invention provides a moving object detection method based on multi-threshold self-optimization background modeling. According to the method, artifacts in a detection result can be quickly eliminated, and the anti-interference capability to a complex environment is relatively high. In order to achieve the purpose, the adopted technical scheme is as follows: creating a background model by using20 neighborhoods of pixel points; judging whether the pixel value of the new frame belongs to the background or not through the simultaneous action of the self-adaptive distance threshold value of thegray scale space and the color distortion threshold value of the RGB space; removing noise points and filling holes in the foreground area through noise point removal, hole filling and median filtering processing; counting the occurrence frequency of foreground points, updating the background model according to the occurrence frequency, wherein a moving target and a background can be displayed onan image through binary segmentation. The method can be widely applied to the fields of traffic video monitoring, indoor security and protection, computer vision and the like, and has a wide application prospect and a high economic value.

Description

technical field

[0001] The invention belongs to the field of image processing, and relates to a moving target detection method, in particular to a moving target detection method based on multi-threshold self-optimizing background modeling. Background technique

[0002] Moving object detection technology is a key technology in the field of computer vision. Its main purpose is to separate the moving objects in the video information from the background, so as to extract clear and complete moving objects. Currently common moving target detection methods include frame difference method, background difference method, mixed Gaussian modeling method, codebook method and visual background extraction method, etc. Among them, the visual background extraction algorithm is a moving object detection algorithm based on random background pixel modeling proposed by Barnich et al. in 2009. It occupies less memory and runs fast, and is suitable for video monitoring and automatic processing fi...

Examples

Embodiment Construction

[0059] The technical solutions of the present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0060] Step 1, build a background model. In order to improve the model quality and avoid repeated selection of pixels, the present invention adopts the modeling of 20 neighborhood pixels of the previous f frame images, and the specific implementation method is as follows:

[0061] Step 101: convert the input image from the RGB space into a grayscale image, the conversion formula is as follows:

[0062] v(x)=0.2989*R+0.5870*G+0.1140*B (1)

[0063] Where v(x) represents the grayscale pixel value converted from the original RGB color space at position x.

[0064] Step 102: Initialize the background model by using the first f frames converted into a grayscale image. It is more appropriate to select 5 for f after many experiments. The expression of the background model M(x) is as follows:

[0065] M(x)={v 1 ,v 2 ,...,v N} (2...

technical field

[0001] The invention belongs to the field of image processing, and relates to a moving target detection method, in particular to a moving target detection method based on multi-threshold self-optimizing background modeling. Background technique

[0002] Moving object detection technology is a key technology in the field of computer vision. Its main purpose is to separate the moving objects in the video information from the background, so as to extract clear and complete moving objects. Currently common moving target detection methods include frame difference method, background difference method, mixed Gaussian modeling method, codebook method and visual background extraction method, etc. Among them, the visual background extraction algorithm is a moving object detection algorithm based on random background pixel modeling proposed by Barnich et al. in 2009. It occupies less memory and runs fast, and is suitable for video monitoring and automatic processing fi...

Examples

Embodiment Construction

[0059] The technical solutions of the present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0060] Step 1, build a background model. In order to improve the model quality and avoid repeated selection of pixels, the present invention adopts the modeling of 20 neighborhood pixels of the previous f frame images, and the specific implementation method is as follows:

[0061] Step 101: convert the input image from the RGB space into a grayscale image, the conversion formula is as follows:

[0062] v(x)=0.2989*R+0.5870*G+0.1140*B (1)

[0063] Where v(x) represents the grayscale pixel value converted from the original RGB color space at position x.

[0064] Step 102: Initialize the background model by using the first f frames converted into a grayscale image. It is more appropriate to select 5 for f after many experiments. The expression of the background model M(x) is as follows:

[0065] M(x)={v 1 ,v 2 ,...,v N} (2...

Images