Moving object detection method based on RPCA and three-frame difference fusion
A three-frame difference, moving target technology, applied in image analysis, image data processing, instruments, etc., can solve a large number of "holes" and other problems, achieve completeness and accuracy, and avoid false detection.
Inactive Publication Date: 2017-08-01
SHAANXI UNIV OF SCI & TECH
View PDF6 Cites 3 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0007] In order to overcome the problem that a large number of "holes" are prone to appear when the traditional three-frame difference method is used for moving target detection, considering that the three-frame difference method is relatively simple and easy to implement in hardware, the present invention aims to provide a motion detection system based on RPCA and three-frame difference fusion. The target detection method adds the video background to the inter-frame difference, thereby avoiding the influence of background pixels on the foreground detection, and can extract a complete and accurate moving foreground in a complex background environment
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0034] The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with examples.
[0035] Such as figure 1 Shown: A moving target detection method based on RPCA and three-frame differential fusion, including the following steps:
[0036] Step1, video data reading: read the input surveillance video as a frame-by-frame picture and save it;
[0037] When the video is processed, the video is first read as a frame of picture I 1 , I 2 ,…I k ∈ N m*n , where m, n are the size of each frame of pictures;
[0038] For the detection of the moving foreground in the kth frame, select the k-1th frame and the kth frame image I k-1 , I k ;
[0039] Step2, grayscale processing: read the current frame image of the video to be processed and convert it into a grayscale image;
[0040] to me k-1 , I k Perform grayscale processing to obtain a grayscale image, denoted as I 1 , I 3 ;
[0041] Step3, RPCA background extraction: ...
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
Login to View More Abstract
The invention provides a moving object detection method based on RPCA and three-frame difference fusion. The method comprises the steps that an input monitoring video is read as pictures of frames and is saved; the current frame picture of the video to be processed is read and converted into a grey-scale image; background extraction is carried out on the current frame of the read video through a RPCA method; the extracted background of the current frame of the video is used as the intermediate frame of three-frame difference; the previous frame of the video is used as the previous frame of three-frame difference; the current frame of the video is used as the last frame of three-frame difference to carry out adjacent difference; and the acquired difference result is thresholded and is subjected to AND to acquire the final detection foreground, wherein the threshold is acquired through repeated experiment and comparison. According to the invention, the complete and accurate movement foreground can be extracted in a complex background environment.
Description
technical field [0001] The invention relates to the technical fields of machine vision and digital image processing, in particular to a moving target detection method based on RPCA and three-frame differential fusion. Background technique [0002] The application of intelligent video surveillance technology will greatly enhance the value of security monitoring in maintaining social public safety. Abnormal human behavior detection in surveillance video is an important part of the intelligent surveillance system, which can give early warning of abnormal behavior in surveillance video. The detection of moving objects in video is the basis of abnormal human behavior detection. Therefore, the research on detection methods of moving objects in videos has very important practical significance. [0003] Traditional moving object detection algorithms include frame difference method, optical flow method and background subtraction method. In the late 1970s, Jain et al. proposed the m...
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
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06T7/20
CPCG06T7/20G06T2207/10016G06T2207/20224
Inventor 亢洁李晓静李静
Owner SHAANXI UNIV OF SCI & TECH