CN106898014A - A kind of intrusion detection method based on depth camera - Google Patents
A kind of intrusion detection method based on depth camera Download PDFInfo
- Publication number
- CN106898014A CN106898014A CN201710098065.8A CN201710098065A CN106898014A CN 106898014 A CN106898014 A CN 106898014A CN 201710098065 A CN201710098065 A CN 201710098065A CN 106898014 A CN106898014 A CN 106898014A
- Authority
- CN
- China
- Prior art keywords
- num0
- thr1
- thr2
- state
- numi
- 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.)
- Granted
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Image Analysis (AREA)
- Burglar Alarm Systems (AREA)
- Closed-Circuit Television Systems (AREA)
Abstract
The present invention provides a kind of intrusion detection method based on depth camera, is detected as follows：User sets monitoring range in camera range ability first；Background pixel number in the default monitoring range of initialization, is designated as num0；Intrusion status are designated as state, state=0 represents nobody, and state=1 represents someone, is initialized as 0；To each frame pitch from view data, pixel quantity in the range of pre-set space is counted, be designated as numi；If numi ＜ num0+thr1, following operation is performed：Num0=(A*num0+B*numi)/(A+B), thr1 are the first thresholds for judging, reflection is background pixel number noise fluctuations situation, and A and B is used to adjust the renewal speed of background pixel number；If numi ＞ num0+thr2, following operation is performed：State is set to 1；Wherein thr2 is the Second Threshold for judging, the erroneous judgement that toy is invaded can be excluded by adjusting thr2；Otherwise State keeps constant.When having invasion in default monitoring range, points can change, and the situation according to change is made whether the judgement that someone invades.Compared with the intrusion detection method of current main flow, the method has not to be influenceed by ambient light substantially, recognition accuracy advantage high.
Description
Technical field
The present invention relates to field of video monitoring, more particularly to a kind of intrusion detection method based on depth camera.
Background technology
Video monitoring is the important component of safety approach system, in recent years, at computer, network and image
Reason, the progress of transmission technology, Video Supervision Technique there has also been significant progress.Intrusion detection is the important set of intelligent video monitoring
Into part, it is widely used with civil area military.Current intrusion detection mainly has based on image/video analysis or infrared
The methods such as detector, wherein the method based on image/video analysis is easily applied by ambient light change, it is more multiple in ambient light
Miscellaneous scene easily produces wrong report；And the method for being based on Infrared Detectors is easily produced by environmental factors such as toy, fog, sleet
Wrong report, additionally mounted mode, angle, the difference of position will also result in certain influence.
The content of the invention
The invention discloses a kind of intrusion detection method based on depth camera, own in obtaining scene by depth camera
, to the range information of camera, and then the point cloud that can obtain scene by the other specification of camera is (every under space coordinates for object
The three-dimensional coordinate set of individual sampled point) data, calculate pixel number in default space monitoring range and go forward side by side Mobile state renewal, when pre-
If points can change when having invasion in monitoring range, the situation according to change is made whether the judgement that someone invades
Its technical scheme is as follows：
A kind of intrusion detection method based on depth camera, it is characterised in that：
Comprise the following steps：
1) configuration monitoring scope：User sets monitoring range in camera range ability；
2) the background pixel number in default monitoring range is initialized, initialization value could be arranged to the corresponding picture of entire image
Prime number (range image of such as 320*240 resolution ratio is set to 320*240=76800) or the frame of power-up initializing first are corresponding pre-
If pixel count in monitoring range or other be more than the numerical value for estimating points in monitoring range, be designated as num0；
3) intrusion status are designated as state, state=0 represents nobody, and state=1 represents someone, is initialized as 0；
4) to each frame pitch from view data, pixel quantity in the range of pre-set space is counted, is designated as numi；
If 5) numi<Num0+thr1, performs following operation：Num0=(A*num0+B*numi)/(A+B),
State is set to 0；Wherein thr1, A, B are customized parameter, thr1 is the first threshold for judging, reflection is
Background pixel number noise fluctuations situation, A and B is used to adjust the renewal speed of background pixel number；
If 6) numi ＞ num0+thr2, following operation is performed：State is set to 1；Wherein thr2 is the second threshold for judging
Value, is customized parameter, and the erroneous judgement of the invasions such as toy can be excluded by adjusting thr2；
If 7) numi is both unsatisfactory for<Num0+thr1, ＞ num0+thr2 are also unsatisfactory for, perform following operation：State keeps
It is constant；
Further, the step 5) in thr1 determine by the following method：
5-1) crawl n times range data (such as N=100) continuous to fixed scene；
Standard deviation (mean square deviation) σ of n times data 5-2) is counted, is calculated according to following criterion calculation formula：
Assuming that there is one group of numerical value X1, X2, X3... Xn (is all real number), and its average value (arithmetic mean of instantaneous value) is μ, formula
Such as schema
Standard deviation is also referred to as standard deviation, or experimental standard deviation, and formula is
Thr1 5-3) is set to certain multiple of σ, thr1 can be generally set to 4 times of σ according to normal distribution, its
In each multiple corresponding area it is as follows：
The area of ■ 95.449974% is in the range of two σ of standard deviation 2 of average value or so；
The area of ■ 99.730020% is in the range of three σ of standard deviation 3 of average value or so；
The area of ■ 99.993666% is in the range of four σ of standard deviation 4 of average value or so.
Further, the step 6) in thr2 determine by the following method：
The spatial volume computing formula that 6-1) each sampled point is represented as
Volume L=(d/f) ^3*A, wherein d is distance, and f is camera focus, and A is the coefficient relevant with camera sensor；
6-2) assume that the object volume for thinking detection is greater than M, then thr2 may be configured as the fixed multiple of M/L, can generally set
0.5 times of M/L is set to, but the value is at least greater than thr1；
Beneficial effect：The present invention provides a kind of intrusion detection method based on depth camera, and depth camera is obtained in scene
All objects and then can obtain the point cloud (space coordinates of scene to the range information of camera by the other specification of camera
The three-dimensional coordinate set of lower each sampled point) data, calculate pixel number in default space monitoring range and go forward side by side Mobile state renewal,
When having invasion in default monitoring range, points can change, and the situation according to change is made whether the judgement that someone invades.
Compared with the intrusion detection method of current main flow, the method has not to be influenceed by ambient light substantially, recognition accuracy advantage high.
Brief description of the drawings
The present invention is described in further detail with reference to accompanying drawing：
Fig. 1 is the application method flow schematic diagram figure；
Specific embodiment
Below by preferred embodiment shown with reference to the accompanying drawings, the present invention is explained in detail, but the invention is not restricted to
The embodiment.
With reference to accompanying drawing 1, the technical program is introduced：
A kind of intrusion detection method based on depth camera, it is characterised in that：
Comprise the following steps：
1) configuration monitoring scope：User sets monitoring range in camera range ability；
2) the background pixel number in default monitoring range is initialized, initialization value could be arranged to the corresponding picture of entire image
Prime number (range image of such as 320*240 resolution ratio is set to 320*240=76800) or the frame of power-up initializing first are corresponding pre-
If pixel count in monitoring range or other be more than the numerical value for estimating points in monitoring range, be designated as num0；
3) intrusion status are designated as state, state=0 represents nobody, and state=1 represents someone, is initialized as 0；
4) to each frame pitch from view data, pixel quantity in the range of pre-set space is counted, is designated as numi；
If 5) numi<Num0+thr1, performs following operation：Num0=(A*num0+B*numi)/(A+B),
State is set to 0；Wherein thr1, A, B are customized parameter, thr1 is the first threshold for judging, reflection is
Background pixel number noise fluctuations situation, A and B is used to adjust the renewal speed of background pixel number；
If 6) numi ＞ num0+thr2, following operation is performed：State is set to 1；Wherein thr2 is the second threshold for judging
Value, is customized parameter, and the erroneous judgement of the invasions such as toy can be excluded by adjusting thr2；
If 7) numi had both been unsatisfactory for ＜ num0+thr1, or has been unsatisfactory for ＞ num0+thr2, following operation is performed：State is protected
Hold constant.
Further, the step 5) in thr1 determine by the following method：
5-1) crawl n times range data (such as N=100) continuous to fixed scene；
Standard deviation (mean square deviation) σ of n times data 5-2) is counted, is calculated according to following criterion calculation formula：
Assuming that there is one group of numerical value X1, X2, X3... Xn (is all real number), and its average value (arithmetic mean of instantaneous value) is μ, formula
Such as schema
Standard deviation is also referred to as standard deviation, or experimental standard deviation, and formula is
Thr1 5-3) is set to certain multiple of σ, thr1 can be generally set to 4 times of σ according to normal distribution, its
In each multiple corresponding area it is as follows：
The area of ■ 95.449974% is in the range of two σ of standard deviation 2 of average value or so；
The area of ■ 99.730020% is in the range of three σ of standard deviation 3 of average value or so；
The area of ■ 99.993666% is in the range of four σ of standard deviation 4 of average value or so.
Further, the step 6) in thr2 determine by the following method：
The spatial volume computing formula that 6-1) each sampled point is represented as
Volume L=(d/f) ^3*A, wherein d is distance, and f is camera focus, and A is the coefficient relevant with camera sensor；
6-2) assume that the object volume for thinking detection is greater than M, then thr2 may be configured as the fixed multiple of M/L, can generally set
0.5 times of M/L is set to, but the value is at least greater than thr1；
The present invention has following features：
1 present invention is performed intrusion detection based on depth camera, and its space coordinates precision for obtaining and data volume are all much high
In common Infrared Detectors, the algorithm provided with reference to the present invention can realize highly stable accurate intrusion detection；
2 present invention devise the background update method of uniqueness, the non-sensor by the change for calculating space object coordinate
The primary light illuminance information of acquisition, can so exclude the interference of ambient light change, change by brightness with common camera
The method for carrying out context update is compared, and more preferably, it is more accurate to update for the method robustness that the present invention is provided；
3 present invention devise the side of the judgement that intrusion object size is carried out using the sampling number obtained in monitoring range
Method, can be very good to exclude the interference (such as toy) of small size object invasion.
Above specific embodiment is merely illustrative of the technical solution of the present invention and unrestricted, although with reference to example to this hair
It is bright to be described in detail, it will be understood by those within the art that, technical scheme can be modified
Or equivalent, without deviating from the spirit and scope of technical solution of the present invention, it all should cover in claim of the invention
In the middle of scope.
Claims (5)
1. a kind of intrusion detection method based on depth camera, it is characterised in that：
Comprise the following steps：
1) configuration monitoring scope：User sets monitoring range in camera range ability；
2) the background pixel number in default monitoring range is initialized：Initialization value may be configured as the corresponding pixel count of entire image or
Pixel count in the corresponding default monitoring range of the frame of power-up initializing first or other be more than the number for estimating points in monitoring range
Value, is designated as num0；
3) intrusion status are designated as state, state=0 represents nobody, and state=1 represents someone, is initialized as 0；
4) to each frame pitch from view data, pixel quantity in the range of pre-set space is counted, is designated as numi；
If 5) numi ＜ num0+thr1, following operation is performed：Num0=(A*num0+B*numi)/(A+B),
State is set to 0；Wherein thr1, A, B are customized parameter, and thr1 is the first threshold for judging, reflection is background
Pixel count noise fluctuations situation, A and B is used to adjust the renewal speed of background pixel number；
If 6) numi ＞ num0+thr2, following operation is performed：State is set to 1；Wherein thr2 is the Second Threshold for judging, is
Customized parameter, the erroneous judgement that toy is invaded can be excluded by adjusting thr2；
If 7) numi is both unsatisfactory for<Num0+thr1, ＞ num0+thr2 are also unsatisfactory for, perform following operation：State keeps not
Become.
2. a kind of intrusion detection method based on depth camera according to claim 1, it is characterised in that：The step 5)
In thr1 determine by the following method：
5-1) crawl n times range data continuous to fixed scene；
The standard deviation sigma of n times data 5-2) is counted, is calculated according to following criterion calculation formula：
Assuming that there is one group of numerical value X1, X2, X3... Xn, its average value is μ, and formula is
Thr1 5-3) is set to certain multiple of σ.
3. a kind of intrusion detection method based on depth camera according to claim 2, it is characterised in that：N is set to
100, thr1 is set to 4 times of σ.
4. a kind of intrusion detection method based on depth camera according to claim 1, it is characterised in that：The step 6)
In thr2 determine by the following method：
The spatial volume computing formula that 6-1) each sampled point is represented as：Volume L=(d/f) ^3*A, wherein d is distance, and f is phase
Machine focal length, A is the coefficient relevant with camera sensor；
6-2) assume that the object volume for thinking detection is greater than M, then thr2 may be configured as the fixed multiple of M/L.
5. a kind of intrusion detection method based on depth camera according to claim 4, it is characterised in that：Thr2 is set to
0.5 times of M/L, and at least it is greater than thr1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710098065.8A CN106898014B (en) | 2017-02-22 | 2017-02-22 | Intrusion detection method based on depth camera |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710098065.8A CN106898014B (en) | 2017-02-22 | 2017-02-22 | Intrusion detection method based on depth camera |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106898014A true CN106898014A (en) | 2017-06-27 |
| CN106898014B CN106898014B (en) | 2020-02-18 |
Family
ID=59185230
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710098065.8A Active CN106898014B (en) | 2017-02-22 | 2017-02-22 | Intrusion detection method based on depth camera |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106898014B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110581980A (en) * | 2018-06-11 | 2019-12-17 | 视锐光科技股份有限公司 | Operation mode of safety monitoring system |
| CN112869462A (en) * | 2020-11-30 | 2021-06-01 | 深圳市博云慧科技有限公司 | Folding bed cabinet |
| CN117173643A (en) * | 2023-11-03 | 2023-12-05 | 深圳市湾测技术有限公司 | Monitoring protection method and device based on 3D camera and related equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102957691A (en) * | 2012-10-12 | 2013-03-06 | 哈尔滨工业大学深圳研究生院 | Cloud intrusion statistical detection method |
| CN103034991A (en) * | 2011-09-29 | 2013-04-10 | 联发科技（新加坡）私人有限公司 | Method and apparatus for foreground object detection |
| CN104424649A (en) * | 2013-08-21 | 2015-03-18 | 株式会社理光 | Method and system for detecting moving object |
| CN104753920A (en) * | 2015-03-01 | 2015-07-01 | 江西科技学院 | Quantum genetic algorithm based intrusion detection method |
-
2017
- 2017-02-22 CN CN201710098065.8A patent/CN106898014B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103034991A (en) * | 2011-09-29 | 2013-04-10 | 联发科技（新加坡）私人有限公司 | Method and apparatus for foreground object detection |
| CN102957691A (en) * | 2012-10-12 | 2013-03-06 | 哈尔滨工业大学深圳研究生院 | Cloud intrusion statistical detection method |
| CN104424649A (en) * | 2013-08-21 | 2015-03-18 | 株式会社理光 | Method and system for detecting moving object |
| CN104753920A (en) * | 2015-03-01 | 2015-07-01 | 江西科技学院 | Quantum genetic algorithm based intrusion detection method |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110581980A (en) * | 2018-06-11 | 2019-12-17 | 视锐光科技股份有限公司 | Operation mode of safety monitoring system |
| CN110581980B (en) * | 2018-06-11 | 2021-06-18 | 视锐光科技股份有限公司 | Operation mode of safety monitoring system |
| CN112869462A (en) * | 2020-11-30 | 2021-06-01 | 深圳市博云慧科技有限公司 | Folding bed cabinet |
| CN117173643A (en) * | 2023-11-03 | 2023-12-05 | 深圳市湾测技术有限公司 | Monitoring protection method and device based on 3D camera and related equipment |
| CN117173643B (en) * | 2023-11-03 | 2024-01-30 | 深圳市湾测技术有限公司 | Monitoring protection method and device based on 3D camera and related equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106898014B (en) | 2020-02-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP6863408B2 (en) | Information processing equipment, information processing methods and programs | |
| US7982774B2 (en) | Image processing apparatus and image processing method | |
| US20090175500A1 (en) | Object tracking apparatus | |
| Shafie et al. | Motion detection techniques using optical flow | |
| CN103735269B (en) | A kind of height measurement method followed the tracks of based on video multi-target | |
| CN107079093B (en) | Calibration device | |
| Liu et al. | On directional k-coverage analysis of randomly deployed camera sensor networks | |
| CN110598590A (en) | Close interaction human body posture estimation method and device based on multi-view camera | |
| CN102982537B (en) | A kind of method and system detecting scene change | |
| CN101216304A (en) | Systems and methods for object dimension estimation | |
| KR101125233B1 (en) | Fusion technology-based security method and security system thereof | |
| CN104240264A (en) | Height detection method and device for moving object | |
| CN106447730A (en) | Parameter estimation method, parameter estimation apparatus and electronic equipment | |
| CN103391430B (en) | DSP (digital signal processor) based relevant tracking method and special device | |
| CN109063625A (en) | A kind of face critical point detection method based on cascade deep network | |
| JP2014006586A (en) | Information processor, and control method and computer program thereof | |
| CN113358231B (en) | Infrared temperature measurement method, device and equipment | |
| CN106898014A (en) | A kind of intrusion detection method based on depth camera | |
| CN112541403B (en) | Indoor personnel falling detection method by utilizing infrared camera | |
| CN109509213B (en) | Harris corner detection method applied to asynchronous time domain vision sensor | |
| CN112184757A (en) | Method and device for determining motion trail, storage medium and electronic device | |
| CN105469054B (en) | The model building method of normal behaviour and the detection method of abnormal behaviour | |
| KR20180096101A (en) | Apparatus and Method for Intelligent Infrared Image Fusion | |
| CN106780539B (en) | Robot vision tracking | |
| CN106600628A (en) | Target object identification method and device based on infrared thermal imaging system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |