CN106056623A - Fixed scene moving object extraction method and device - Google Patents

Fixed scene moving object extraction method and device Download PDF

Info

Publication number
CN106056623A
CN106056623A CN201610334633.5A CN201610334633A CN106056623A CN 106056623 A CN106056623 A CN 106056623A CN 201610334633 A CN201610334633 A CN 201610334633A CN 106056623 A CN106056623 A CN 106056623A
Authority
CN
China
Prior art keywords
image
difference
moving object
images
fixed scene
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201610334633.5A
Other languages
Chinese (zh)
Inventor
贺永刚
邹超洋
万美君
朱豪
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nanchang Black Shark Technology Co Ltd
Original Assignee
Shenzhen Zeusis Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Shenzhen Zeusis Technology Co Ltd filed Critical Shenzhen Zeusis Technology Co Ltd
Priority to CN201610334633.5A priority Critical patent/CN106056623A/en
Publication of CN106056623A publication Critical patent/CN106056623A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10016—Video; Image sequence
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10024—Color image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20212—Image combination
    • G06T2207/20224—Image subtraction

Landscapes

  • Image Analysis (AREA)

Abstract

The invention provides a fixed scene moving object extraction method and device. The method comprises the steps of: S101, obtaining and marking N frames of images in a video stream; S102, setting that the ith frame is an object image, subtracting the ith frame of image from other images respectively, calculating the differences, taking the absolute values, and obtaining a group of difference images; S103, converting the difference images into binary images through a threshold; S104, for all images, accumulating all corresponding position elements and obtaining one accumulated image; S105, carrying out screening on the values of the accumulated image, and selecting all the matrix image blocks whose value is N-1, namely moving objects. The sampling frame number is adjusted, so that the moving object extraction effect and efficiency of the algorithm are balanced, and the algorithm is enabled to applicable to various hardware platforms. The algorithm is free from the influences of sampling frame intervals and sequences, so that the algorithm is more flexible. A relatively large image frame number is utilized to avoid the condition that the objects are not staggered among all frames, so that the integrity of the obtained moving objects is greatly improved.

Description

Fixed scene Extracting of Moving Object and device
Technical field
The present embodiments relate to video capture technology, particularly relate to extracting method and the device of video frequency motion target
Background technology
At present, from fixed scene, focus and the difficult point that moving target is computer vision research is extracted.In prior art, Utilize the front N frame image data of video to background constructing model to build background image, by being made with background image by present frame Difference detects the moving target in present image.The method needs substantial amounts of frame of video to be built up by the model of background Come.Additionally, when the first frame has operational objective, background modeling can be using moving target as background so that the fortune of the most some frames Moving-target accuracy of detection declines.
Frame difference method is the another kind of method of detection moving target.The frame difference individually calculated between video can make target before motion After position obvious moving region all occurs, it is impossible to accurately estimate moving target.Patent CN103020991B is according to inciting somebody to action Frame difference method combines with background modeling method and estimates the position of moving target.The combination of two kinds of methods reduces motion virtually The detection efficiency of target.Patent CN102881025A is by the catch cropping difference of frame before and after present frame and its, between error image With operation obtain moving target.The method, when extracting moving target, is affected by sampling frame number and interframe order, it is impossible to Well extract moving target, need the follow-up morphologic filtering etc. that carries out to operate.
In sum, in order to solve the problems referred to above, the method that this patent proposes to utilize frame difference two-value to accumulate is to estimate image In moving target.
Summary of the invention
It is an object of the invention to provide a kind of fixed scene Extracting of Moving Object and device, for reliably and effectively Moving target is extracted in fixed scene.
The present invention provides a kind of fixed scene Extracting of Moving Object, including: S101: obtain N frame figure in video streaming Picture, and labelling；S102: set the i-th frame as target image to be estimated, by other images and the i-th two field picture subtraction calculations difference And take absolute value, obtain one group of error image；S103: be bianry image by a threshold transitions by described error image；S104: A width accumulated image is obtained for the summation that adds up of all image correspondence position elements；S105: the value on described assignment graph picture is entered Row filter, selects the matrix image block that all values is N-1 and is moving target.
Preferably, in described S101, described N two field picture can obtain with continuous acquisition or interval gathers and obtains.
Preferably, in described S102, described difference is the data in different colours space.
Preferably, in described S102, described difference is gray level image.
Preferably, in described S102, described difference is the norm distance between image pixel feature.
The present invention also provides for a kind of device implementing above-mentioned fixed scene Extracting of Moving Object, including: image acquisition Unit, obtains N two field picture, and labelling in video streaming；Image pre-processing unit, sets the i-th frame as target image, by other figures As with the i-th two field picture subtraction calculations difference taking absolute value, obtain one group of error image；Image computing unit, by described difference Image is bianry image by a threshold transitions；Graphics processing unit, sues for peace for all image correspondence position elements are cumulative To a width accumulated image；Objective extraction unit, screens the value in described accumulated image, selects the square that all values is N-1 Battle array image block is moving target.
Preferably, N two field picture described in described image acquisition units can obtain with continuous acquisition or interval gathers and obtains.
Preferably, difference described in described image pre-processing unit can be the data in different colours space.
Preferably, difference described in described image pre-processing unit can be gray level image.
Preferably, difference described in described image pre-processing unit can be the norm distance between image pixel feature.
The invention has the beneficial effects as follows, by regulation sampling frame number so that algorithm is extracting effect and the effect of moving target Rate aspect is balanced so that algorithm can be applicable to multiple hardwares platform.Do not affected by sample frame interval and order, made Obtain the most flexible of algorithm.More number of image frames can be utilized to avoid the occurrence of the target situation in all interframe dislocation-free, greatly Promote the integrity of moving target obtained.
Accompanying drawing explanation
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing In having technology to describe, the required accompanying drawing used is briefly described, it should be apparent that, the accompanying drawing in describing below is this Some bright embodiments, for those of ordinary skill in the art, on the premise of not paying creative work, it is also possible to Other accompanying drawing is obtained according to these accompanying drawings.
Fig. 1 is fixed scene Extracting of Moving Object schematic flow sheet of the present invention；
Fig. 2 is fixed scene moving target recognition device schematic diagram of the present invention.
Reference:
S101～S105 step
201 image acquisition units 202 image pre-processing unit 203 image computing units
204 graphics processing unit 205 Objective extraction unit
Detailed description of the invention
For making the purpose of the embodiment of the present invention, technical scheme and advantage clearer, below in conjunction with the embodiment of the present invention In accompanying drawing, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment is The a part of embodiment of the present invention rather than whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art The every other embodiment obtained under not making creative work premise, broadly falls into the scope of protection of the invention.
Fixed scene Extracting of Moving Object schematic flow sheet the most of the present invention, this example is for carrying by intelligence Rise the terminal unit effect when extracting moving target, including: S101: obtain N two field picture, and labelling in video streaming；S102: Set the i-th frame as target image, by other images and the i-th two field picture subtraction calculations difference and take absolute value, obtain one group of difference Image；S103: be bianry image by a threshold transitions by described error image；S104: for all image correspondence positions unit The cumulative summation of element obtains a width accumulated image；S105: screen the value in described accumulated image, selecting all values is N-1 Matrix image block be moving target.
Wherein, S101 obtains N two field picture in video streaming, and view data is concentrated in addition to present frame, at least to have 2 frame figures Picture.The collection of view data can be with acquisition order, it is also possible to random choose.More number of image frames can be extracted and transport the most accurately Moving-target, but operation efficiency can decline.Can be compromised in arithmetic accuracy and efficiency by regulation frame number.Ability Field technique personnel understand, as long as N can realize the purpose of the present invention more than or equal to 2, reach the technique effect of the present invention, fall into Protection scope of the present invention.
Further, in the present embodiment, the method calculating difference in S102 has multiple.Several calculating difference given below Method, as reference, skilled person will appreciate that, any non-deformation obtained by creative work is all the guarantor of the present invention In the range of protecting.
The first, calculate difference and can use the data in different colours space, such as, uses RGB color:
Dij=| Ri-Rj|+|Gi-Gj|+|Bi-Bj|
Wherein, RiRepresent image IiR component data, GiRepresent image IiG component data, BiRepresent image IiB divide Amount data.
The second, calculates difference and can use gray level image
Dij=| Si-Sj|
Wherein, SiRepresent the i-th width image IiGray level image, Si=(Ri+Gi+Bi)/3。
The third, calculating difference can be the norm distance between image pixel feature.
Dij(x, y)=| | Fi(x,y)-Fj(x,y)||2
Wherein, | |. | |2For L2 norm distance, Fi(x is y) at image IiOn a characteristic vector.Wherein, feature to The computational methods of amount also have a lot, such as calculate point (x, y) rectangular histogram of position one sub regions.
Further in preferred embodiment, in S103, by threshold value difference data is converted into two-value data:
Wherein, Dij(x, y) be the i-th two field picture with jth two field picture difference after (x, y) value at coordinate points, the B of imageij For corresponding bianry image, thr is threshold value set in advance.
In another embodiment of the present invention, S104 is specially and accumulates all binary image datas:
A i = Σ j = 1 j ≠ i N B i j .
For in S105, by above-mentioned accumulative image Ai, being worth the data for N-1 is prospect, and remaining is background, therefore i-th The movement destination image of two field picture is:That is, the image block of all N-1 numerical value is screened Out, the moving target that the present invention needs it is.
An alternative embodiment of the invention provides a kind of device implementing above-mentioned fixed scene Extracting of Moving Object, bag Include: image acquisition units 201, obtain N two field picture, and labelling in video streaming；Image pre-processing unit 202, set the i-th frame as Target image, by other images and the i-th two field picture subtraction calculations difference and take absolute value, obtains one group of error image；Image meter Calculate unit 203, be bianry image by described error image by a threshold transitions；Graphics processing unit 204, for all images The summation that adds up of correspondence position element obtains an accumulated image；Objective extraction unit 205, sieves the value on described assignment graph picture Choosing, selects the matrix image block that all values is N-1 and is moving target.
Wherein, image acquisition units 201 obtains N two field picture in video streaming, and view data is concentrated in addition to present frame, at least There are 2 two field pictures.The collection of view data can be with acquisition order, it is also possible to random choose.More number of image frames can be extracted more For accurate moving target, but operation efficiency can decline.Can be enterprising in arithmetic accuracy and efficiency by regulation frame number Row compromise.Skilled person will appreciate that, as long as N can realize the purpose of the present invention more than or equal to 2, reach the skill of the present invention Art effect, falls into protection scope of the present invention.
Further, the method calculating difference in image pre-processing unit 202 can have multiple, specifically refers to hereinbefore That mentions can use the data in different colours space, gray level image can be used can be maybe the norm between image pixel feature Distance.Specific algorithm repeats no more.
Separately, previously mentioned further algorithm in steps, can realize in the corresponding unit in this device.Tool Body is, image computing unit 203 can be converted into two-value data by threshold value to difference data.Graphics processing unit 204 is specifically used In specially all binary image datas being accumulated as image Ai.Objective extraction unit 205 is specifically for by above-mentioned accumulative image Ai In, being worth the data for N-1 is prospect, and remaining is background, obtains the moving image of the i-th frame.Specific algorithm repeats no more.
The present invention is by regulation sampling frame number so that algorithm is put down in terms of the effect extracting moving target and efficiency Weighing apparatus so that algorithm can be applicable to multiple hardwares platform.Do not affected by sample frame interval and order so that algorithm is more Flexibly.More number of image frames can be utilized to avoid the occurrence of the target situation in all interframe dislocation-free, promote acquisition greatly The integrity of moving target.
Last it is noted that various embodiments above is only in order to illustrate technical scheme, it is not intended to limit；To the greatest extent The present invention has been described in detail by pipe with reference to foregoing embodiments, it will be understood by those within the art that: it depends on So the technical scheme described in foregoing embodiments can be modified, or the most some or all of technical characteristic is entered Row equivalent；And these amendments or replacement, do not make the essence of appropriate technical solution depart from various embodiments of the present invention technology The scope of scheme.

Claims (10)

1. a fixed scene Extracting of Moving Object, it is characterised in that including:
S101: obtain N two field picture, and labelling in video streaming；
S102: set the i-th frame as target image to be calculated, by other images and the i-th two field picture subtraction calculations difference and take absolutely To value, obtain one group of error image；
S103: be bianry image by a threshold transitions by described error image；
S104: obtain a width accumulated image for the summation that adds up of all bianry image correspondence position elements；
S105: screen the value in described accumulated image, selects the matrix image block that all values is N-1 and is motion mesh Mark.
Fixed scene Extracting of Moving Object the most according to claim 1, it is characterised in that in described S101, described N Two field picture can obtain with continuous acquisition or interval gathers and obtains.
Fixed scene Extracting of Moving Object the most according to claim 1, it is characterised in that in described S102, described Difference is the data in different colours space.
Fixed scene Extracting of Moving Object the most according to claim 1, it is characterised in that in described S102, described Difference is gray level image.
Fixed scene Extracting of Moving Object the most according to claim 1, it is characterised in that in described S102, described Difference is the norm distance between image pixel feature.
6. the device of the fixed scene Extracting of Moving Object that a kind is implemented described in claim 1, it is characterised in that including:
Image acquisition units, obtains N two field picture, and labelling in video streaming；
Image pre-processing unit, sets the i-th frame as target image, by other images and the i-th two field picture subtraction calculations difference and take Absolute value, obtains one group of error image；
Image computing unit, is bianry image by described error image by a threshold transitions；
Graphics processing unit, obtains a width accumulated image for the summation that adds up of all bianry image correspondence position elements；
Objective extraction unit, screens the value on described assignment graph picture, selects the matrix image block that all values is N-1 and is Moving target.
Extraction element the most according to claim 6, it is characterised in that described in described image acquisition units, N two field picture can Obtain with continuous acquisition or interval gathers and obtains.
Extraction element the most according to claim 6, it is characterised in that described in described image pre-processing unit, difference is permissible Data for different colours space.
Extraction element the most according to claim 6, it is characterised in that described in described image pre-processing unit, difference is permissible For gray level image.
Extraction element the most according to claim 6, it is characterised in that described in described image pre-processing unit, difference can Think the norm distance between image pixel feature.
CN201610334633.5A 2016-05-18 2016-05-18 Fixed scene moving object extraction method and device Pending CN106056623A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610334633.5A CN106056623A (en) 2016-05-18 2016-05-18 Fixed scene moving object extraction method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610334633.5A CN106056623A (en) 2016-05-18 2016-05-18 Fixed scene moving object extraction method and device

Publications (1)

Publication Number Publication Date
CN106056623A true CN106056623A (en) 2016-10-26

Family

ID=57177183

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610334633.5A Pending CN106056623A (en) 2016-05-18 2016-05-18 Fixed scene moving object extraction method and device

Country Status (1)

Country Link
CN (1) CN106056623A (en)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060039584A1 (en) * 2004-08-23 2006-02-23 Denso Corporation Motion detection method and device, program and vehicle surveillance system
CN102881025A (en) * 2012-09-17 2013-01-16 天津工业大学 Method for detecting multiple moving targets

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060039584A1 (en) * 2004-08-23 2006-02-23 Denso Corporation Motion detection method and device, program and vehicle surveillance system
CN102881025A (en) * 2012-09-17 2013-01-16 天津工业大学 Method for detecting multiple moving targets

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
左凤艳: "基于特征点的运动目标检测和跟踪算法研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 *

Similar Documents

Publication Publication Date Title
Kopsiaftis et al. Vehicle detection and traffic density monitoring from very high resolution satellite video data
CN102368821B (en) Adaptive noise intensity video denoising method and system thereof
CN107256225B (en) Method and device for generating heat map based on video analysis
WO2015117072A1 (en) Systems and methods for detecting and tracking objects in a video stream
CN105184812B (en) A kind of pedestrian based on target following hovers detection method
CN105574894B (en) A kind of screening technique and system of moving object feature point tracking result
CN101996406A (en) No-reference structure definition image quality evaluation method
CN104680483B (en) The noise estimation method of image, video image denoising method and device
CN101299275A (en) Method and device for detecting target as well as monitoring system
CN109359593B (en) Rain and snow environment picture fuzzy monitoring and early warning method based on image local grid
CN106851302B (en) A kind of Moving Objects from Surveillance Video detection method based on intraframe coding compression domain
CN102855465B (en) A kind of tracking of mobile object
CN110114801B (en) Image foreground detection device and method and electronic equipment
CN103258332A (en) Moving object detection method resisting illumination variation
CN106204617A (en) Adapting to image binarization method based on residual image rectangular histogram cyclic shift
CN104717400A (en) Real-time defogging method of monitoring video
CN104658009A (en) Moving-target detection method based on video images
US20170103536A1 (en) Counting apparatus and method for moving objects
CN106056078A (en) Crowd density estimation method based on multi-feature regression ensemble learning
CN109711313B (en) Real-time video monitoring method for identifying sewage poured into river channel
CN103578121A (en) Motion detection method based on shared Gaussian model in disturbed motion environment
CN105427276A (en) Camera detection method based on image local edge characteristics
CN103996199A (en) Movement detection method based on depth information
CN106056623A (en) Fixed scene moving object extraction method and device
CN106251345A (en) A kind of method that real-time monitoring video contrast is abnormal

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
TA01 Transfer of patent application right
TA01 Transfer of patent application right

Effective date of registration: 20180224

Address after: 330008 Jiangxi Province, Nanchang Qingshan Lake Economic and Technological Development Zone of Nanchang Jiao Town Office Building Room 319

Applicant after: Nanchang shark Technology Co. Ltd.

Address before: 518063 Guangdong city of Shenzhen province Nanshan District Wei new software park building 1 floor 2 South

Applicant before: SHENZHEN ZEUSIS TECHNOLOGY CO., LTD.

WD01 Invention patent application deemed withdrawn after publication
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20161026