CN101483001A - Video-based intrusion detection method, detection apparatus and chip - Google Patents

Abstract

The present invention provides an invasion detection method based on video, detection device, chip, wherein, the method includes: establish reference matrix and contrasting detection matrix; and then comparing the reference matrix and contrasting detection matrix; determining the technological means of the invasion object based on the comparison result, thereby overcoming the defect of poor accuracy brought by frame difference detection based on video frame pixel points in the prior art, further automatically removing those sport interference produced by natural reasons and improving the accuracy of intrusion detection.

Description

Intrusion detection method, pick-up unit, chip based on video

Technical field

The present invention relates to field of video monitoring, especially, relate to a kind of intrusion detection method, pick-up unit, chip based on video.

Background technology

Intelligent video monitoring is based on computer vision technique the video image in the monitoring scene is analyzed; extract the key message in the scene; and the monitor mode of formation corresponding event and alarm, application is very widely arranged at aspects such as public safety protection, traffic administrations.

Intelligent monitor system generally all is to use video camera or IP Camera that a certain scene is continuously taken, and the video flowing of collecting is handled, and detects intrusion target, and to intrusion target classify, subsequent treatment such as tracking, identification.The video intrusion detection is the basis of intelligent monitoring, and follow-up processing is had important meaning.

Intrusion target detection method commonly used at present comprises time differencing method and background subtraction point-score.Time differencing method is called the frame-to-frame differences point-score again, and is all constant by the pixel value and the position of pixel in the hypothesis background image, thereby isolates background image and foreground image, and here, so-called " foreground image " is potential invasion object.Time differencing method has multiple implementation method, and wherein a kind of is at continuous video image (also claiming video flowing), perhaps carries out absolute calculus of differences between each two field picture of image sequence, and algorithm flow is, two two field picture f in video flowing or the image sequence _kAnd f _K-1Carry out absolute calculus of differences and obtain difference image, again difference image is carried out thresholding and handle and to obtain binary image, use the method for mathematical morphology that binary image is carried out Filtering Processing then and obtain foreground image R _k, again foreground image is carried out connectivity analysis, for example, fill cavity in the foreground image, take out the less isolated area of area, non-connected region etc., differentiate at last, only keep area in the connected region greater than given threshold value connected component.Thereby isolate foreground image and background image, just found the invasion object.The background subtraction point-score is by current frame image f _kWith with average background image b _K-1Carry out calculus of differences, thereby isolate background image and foreground image.The algorithm flow basically identical of the algorithm flow of background difference and time difference.

This shows; existing Detection for Moving Target mainly is based on the detection technique of pixel; a shortcoming of this method is; this method is very responsive to noise; in constant indoor of scene; can obtain effect relatively preferably; and when scene changes slightly; for example; in actual application, monitored zone takes place because the picture activity that natural cause produces (is waved such as branch through regular meeting; the neon bulb flicker; night the starlight level the figure viewed from behind etc.), extract the intruding object detection precision very poor; detection system thinks it is the object that illegal invasion is arranged by mistake, therefore can produce wrong report.

In view of this, need the urgent technical matters that solves of those skilled in the art to be exactly: how in the process of intrusion detection, to eliminate as much as possible because the motion artifacts clocklike of natural cause generation.

Summary of the invention

Technical matters to be solved by this invention is: a kind of intrusion detection method based on video is provided, based on this method, can eliminates automatically because the motion artifacts clocklike that natural cause produces in the process of intrusion detection.

In order to address the above problem, the invention discloses a kind of video intrusion detection method, described method comprises: preset have identical initial value with reference to matrix and comparison and detection matrix, and described size with reference to matrix and described comparison and detection matrix is corresponding with the resolution of described video; Preset initialization cycle and video intrusion detection cycle; Described method comprises the steps: in described initialization cycle, sets up described with reference to matrix; In the described video intrusion detection cycle, set up described comparison and detection matrix; Compare with reference to matrix and described comparison and detection matrix described, and, determine the invasion object according to comparative result.

Preferably, in described initialization cycle, set up the poor of the described value that further comprises the steps: to calculate corresponding pixel points in the value of each pixel in one two field picture of back and the former frame image with reference to matrix, obtain frame difference matrix; According to described frame difference matrix, obtain the moving surface of a plurality of separation; Calculate the center of described each moving surface, and be provided with according to preset rules with reference to the element corresponding to described center position on the matrix described; Do not have under the overtime situation at described initialization cycle, return and carry out the described step of obtaining frame difference matrix.

Preferably, described initialization cycle is greater than the described video intrusion detection cycle.

Preferably, in the described video intrusion detection cycle, set up described comparison and detection matrix, further comprise the steps: to calculate the poor of corresponding pixel points in the value of each pixel in one two field picture of back and the former frame image, obtain frame difference matrix; According to described frame difference matrix, obtain the moving surface of a plurality of separation; Calculate the center of described each moving surface, and the element corresponding to described center position on the described comparison and detection matrix is provided with according to preset rules; Judge whether the described video intrusion detection cycle is overtime, if not, then return the described step of obtaining frame difference matrix.

Preferably, compare with reference to matrix and described comparison and detection matrix, and, determine the invasion object, further comprise the steps: to preset the variance rate threshold value according to comparative result with described; According to described distribution, determine abnormal area with reference to element in the matrix; In described comparison and detection matrix, search and the corresponding zone of described abnormal area; Add up between described abnormal area and the described corresponding zone difference of element coordinate, and calculated difference rate; Judge that whether described variance rate is greater than described variance rate threshold value, if then this abnormal area has the invasion object; If not, then this abnormal area is not invaded object.

Preferably, described R-matrix initial value is 0, described comparison and detection matrix initial value is 0.

Preferably, the center of described each moving surface of described calculating, and be specially according to preset rules setting with reference to the element corresponding to described center position on the matrix described: the element corresponding to described center position on the described R-matrix is set to 1.

Preferably, described center is the geometric center of moving surface.

According to another embodiment of the present invention, a kind of video invasion detecting device is also disclosed, comprising: the preset parameter unit, set up unit, comparison and detection matrix with reference to matrix and set up unit, invasion object determining unit.

Wherein, the preset parameter unit be used to preset have identical initial value with reference to matrix and comparison and detection matrix, and described size with reference to matrix and described comparison and detection matrix is corresponding with the resolution of described video; Preset initialization cycle and video intrusion detection cycle; Set up the unit with reference to matrix and be used in described initialization cycle, set up described with reference to matrix; The comparison and detection matrix is set up the unit and is used for setting up described comparison and detection matrix in the described video intrusion detection cycle; Invasion object determining unit is used for comparing with reference to matrix and described comparison and detection matrix described, and according to comparative result, determines the invasion object.

Preferably, describedly set up the unit with reference to matrix and further comprise: the first frame difference matrix acquisition module, the first moving surface acquisition module, first are provided with module, first judge module.

Wherein, the first frame difference matrix acquisition module is used for calculating the poor of corresponding pixel points in the value of back one each pixel of two field picture and the former frame image, obtains frame difference matrix; The first moving surface acquisition module is used for obtaining the moving surface of a plurality of separation according to described frame difference matrix; First is provided with the center that module is used to calculate described each moving surface, and is provided with according to preset rules with reference to the element corresponding to described center position on the matrix described; First judge module is used to judge whether described initialization cycle is overtime, if not, and the gating first frame difference matrix acquisition module then.

Preferably, described comparison and detection matrix is set up the unit and further comprised: the second frame difference matrix acquisition module, the second moving surface acquisition module, second are provided with module, second judge module.

Wherein, the second frame difference matrix acquisition module is used for calculating the poor of corresponding pixel points in the value of back one each pixel of two field picture and the former frame image, obtains frame difference matrix; The second moving surface acquisition module is used for obtaining the moving surface of a plurality of separation according to described frame difference matrix; Second is provided with the center that module is used to calculate described each moving surface, and the element corresponding to described center position on the described comparison and detection matrix is provided with according to preset rules; Second judge module is used to judge whether the described video intrusion detection cycle is overtime, if not, and the gating second frame difference matrix acquisition module then.

Preferably, described invasion object determining unit further comprises: module, variance rate computing module, the 3rd judge module are searched in variance rate threshold value preset module, abnormal area determination module, corresponding region.

Wherein, variance rate threshold value preset module is used to preset the variance rate threshold value; The abnormal area determination module is used for determining abnormal area according to described distribution with reference to the matrix element; The corresponding region is searched module and is used at described comparison and detection matrix, searches and the corresponding zone of described abnormal area; The variance rate computing module is used to add up between described abnormal area and the described corresponding zone, the difference of element coordinate, and calculated difference rate; The 3rd judge module is used to judge that whether described variance rate is greater than described variance rate threshold value, if then this abnormal area has the invasion object; If not, then this abnormal area is not invaded object.

Preferably, described R-matrix initial value is 0, described comparison and detection matrix initial value is 0.

Preferably, the center of described each moving surface of described calculating, and with described with reference to being provided with further according to preset rules corresponding to the element of described center position on the matrix be: the element corresponding to described center position on the described R-matrix is set to 1.

Preferably, described center is the geometric center of moving surface.

According to another embodiment of the present invention, a kind of chip that is used for the video intrusion detection is also disclosed, described chip comprises the video invasion detecting device, and described video invasion detecting device comprises: the preset parameter unit, set up unit, comparison and detection matrix with reference to matrix and set up unit, invasion object determining unit.

Wherein, the preset parameter unit be used to preset have identical initial value with reference to matrix and comparison and detection matrix, and described size with reference to matrix and described comparison and detection matrix is corresponding with the resolution of described video; Preset initialization cycle and video intrusion detection cycle; Set up the unit with reference to matrix and be used in described initialization cycle, set up described with reference to matrix; The comparison and detection matrix is set up the unit and is used for setting up described comparison and detection matrix in the described video intrusion detection cycle; Invasion object determining unit is used for comparing with reference to matrix and described comparison and detection matrix described, and according to comparative result, determines the invasion object.

Preferably, describedly set up the unit with reference to matrix and further comprise: the first frame difference matrix acquisition module, the first moving surface acquisition module, first are provided with module, first judge module.

Wherein, the first frame difference matrix acquisition module is used for calculating the poor of corresponding pixel points in the value of back one each pixel of two field picture and the former frame image, obtains frame difference matrix; The first moving surface acquisition module is used for obtaining the moving surface of a plurality of separation according to described frame difference matrix; First is provided with the center that module is used to calculate described each moving surface, and is provided with according to preset rules with reference to the element corresponding to described center position on the matrix described; First judge module is used to judge whether described initialization cycle is overtime, if not, and the described first frame difference matrix acquisition module of gating then.

Preferably, described comparison and detection matrix is set up the unit and further comprised: the second frame difference matrix acquisition module, the second moving surface acquisition module, second are provided with module, second judge module.

Wherein, the second frame difference matrix acquisition module is used for calculating the poor of corresponding pixel points in the value of back one each pixel of two field picture and the former frame image, obtains frame difference matrix; The second moving surface acquisition module is used for obtaining the moving surface of a plurality of separation according to described frame difference matrix; Second is provided with the center that module is used to calculate described each moving surface, and the element corresponding to described center position on the described comparison and detection matrix is provided with according to preset rules; Second judge module is used to judge whether the described video intrusion detection cycle is overtime, if not, and the gating second frame difference matrix acquisition module then.

Preferably.Described invasion object determining unit further comprises: module, variance rate computing module, the 3rd judge module are searched in variance rate threshold value preset module, abnormal area determination module, corresponding region.

Wherein, variance rate threshold value preset module is used to preset the variance rate threshold value; The abnormal area determination module is used for determining abnormal area according to described distribution with reference to the matrix element; The corresponding region is searched module and is used at described comparison and detection matrix, searches and the corresponding zone of described abnormal area; The variance rate computing module is used to add up between described abnormal area and the described corresponding zone, the difference of element coordinate, and calculated difference rate; The 3rd judge module is used to judge that whether described variance rate is greater than described variance rate threshold value, if then this abnormal area has the invasion object; If not, then this abnormal area is not invaded object.

Compared with prior art, the present invention has the following advantages:

Make full use of because natural cause causes the motion clocklike that produces, the principle that its moving surface center is constant, by setting up with reference to matrix and comparison and detection matrix, compare with reference to matrix and described comparison and detection matrix described then, and according to comparative result, determine the technological means of invasion object, overcome in the prior art, detect the defective of the low precision that is brought based on frame of video pixel point frame difference, and then eliminate the motion artifacts clocklike that those produce owing to natural cause automatically, and then improved the precision of intrusion detection.

Description of drawings

Fig. 1 is the flow chart of steps of the video intrusion detection method embodiment according to the present invention;

Fig. 2 among the video intrusion detection method embodiment, in described initialization cycle, sets up described flow chart of steps with reference to matrix according to the present invention;

Fig. 3 among the video intrusion detection method embodiment, in the described video intrusion detection cycle, sets up the flow chart of steps of described ginseng comparison and detection matrix according to the present invention;

Fig. 4, compares with reference to matrix and described comparison and detection matrix described among the video intrusion detection method embodiment according to the present invention, and according to comparative result, determines the flow chart of steps of invasion object;

Fig. 5 is the structural representation of the video detecting device embodiment according to the present invention;

Fig. 6 in the video detecting device, sets up the structural representation of unit, unit according to the present invention with reference to matrix;

Fig. 7 is according to the present invention in the video detecting device, and the comparison and detection matrix is set up the structural representation of unit, unit;

Fig. 8 in the video detecting device, invades the structural representation of object determining unit according to the present invention.

Embodiment

For above-mentioned purpose of the present invention, feature and advantage can be become apparent more, the present invention is further detailed explanation below in conjunction with the drawings and specific embodiments.

Core concept of the present invention is: make full use of because natural cause causes the motion clocklike that produces, the principle that its moving surface center is constant, by setting up with reference to matrix and comparison and detection matrix, compare with reference to matrix and described comparison and detection matrix described then, and according to comparative result, determine the technological means of invasion object, overcome in the prior art, detect the defective of the low precision that is brought based on frame of video pixel point frame difference, and then eliminate the motion artifacts clocklike that those produce owing to natural cause automatically, and then improved the precision of intrusion detection.

With reference to Fig. 1, Fig. 1 is the flow chart of steps of the video intrusion detection method embodiment according to the present invention, comprises the steps:

Step 101: preset have identical initial value with reference to matrix and comparison and detection matrix, and described size with reference to matrix and described comparison and detection matrix is corresponding with the resolution of described video; Preset initialization cycle and video intrusion detection cycle;

Step 102: in described initialization cycle, set up described with reference to matrix;

Step 103: in the described video intrusion detection cycle, set up described comparison and detection matrix;

Step 104: compare with reference to matrix and described comparison and detection matrix described, and, determine the invasion object according to comparative result.

With reference to Fig. 2, Fig. 2 among the video intrusion detection method embodiment, in described initialization cycle, sets up described flow chart of steps with reference to matrix according to the present invention, comprises the steps:

Step 201: calculate value poor of corresponding pixel points in the value of each pixel in one two field picture of back and the former frame image, obtain frame difference matrix;

Step 202:, obtain the moving surface of a plurality of separation according to described frame difference matrix;

Step 203: calculate the center of described each moving surface, and be provided with according to preset rules with reference to the element corresponding to described center position on the matrix with described;

Step 204: judge whether described initialization cycle is overtime, if not, then return step 201, if, execution in step 205;

Step 205, initialization finishes with reference to matrix, and storage is with reference to matrix.

Generally speaking, this process of setting up with reference to matrix is a process of " during length " relatively, and in the application of reality, this is set up process and can be controlled at about 10 minutes.

With reference to Fig. 3, Fig. 3 among the video intrusion detection method embodiment, in the described video intrusion detection cycle, sets up the flow chart of steps of described comparison and detection matrix according to the present invention, comprises the steps:

Step 301: calculate the poor of corresponding pixel points in the value of each pixel in one two field picture of back and the former frame image, obtain frame difference matrix;

Step 302:, obtain the moving surface of a plurality of separation according to described frame difference matrix;

Step 303: calculate the center of described each moving surface, and the element corresponding to described center position on the described comparison and detection matrix is provided with according to preset rules;

Step 304: judge whether the described video intrusion detection cycle is overtime, if not, then return the described step of obtaining frame difference matrix.

Wherein, in step 304, under overtime situation of video intrusion detection cycle, then the comparison and detection matrix that should obtain by step 303 for the last time in the cycle is defined as the comparison and detection matrix in this intrusion detection cycle.

Generally speaking, the process of setting up of this comparison and detection matrix matrix is " in short-term " process, can only utilize several frames, such as 2 frames.

With reference to Fig. 4, Fig. 4, compares with reference to matrix and described comparison and detection matrix described among the video intrusion detection method embodiment according to the present invention, and according to comparative result, determines the flow chart of steps of invasion object, comprises the steps:

Step 401: preset the variance rate threshold value;

Step 402:, determine abnormal area according to described distribution with reference to element in the matrix;

Step 403: in described comparison and detection matrix, search and the corresponding zone of described abnormal area;

Step 404: add up between described abnormal area and the described corresponding zone difference of element coordinate, and calculated difference rate;

Step 405: the size of more described variance rate and described variance rate threshold value; If the former is greater than the latter, then this abnormal area has the invasion object; Otherwise this abnormal area is not invaded object.

Preferably, described R-matrix initial value is 0, described comparison and detection matrix initial value is 0.The benefit of doing like this is, gets up for system handles, and data volume is smaller.Certainly other numerical value also can, the present invention is identical at the initial value that this limits two matrixes, what is as for initial value, the present invention does not do qualification.

Preferably, in step: the center of calculating described each moving surface, and be provided with according to preset rules with reference to the element corresponding to described center position on the matrix described, the rule that is provided with wherein is: the element corresponding to described center position on the described R-matrix is set to 1.In fact, we can be provided with any numerical value outside the junior one, this to application of the present invention can't produce up to influence, 1 can, 2 can, 3 also can ..., being preferably 1 reason is to consider the data volume of system equally, selects 1 the calculated amount of system is diminished.

Preferably, described center is the geometric center of moving surface, and selecting geometric center for use is to make full use of because natural cause causes the motion clocklike that produces, and the principle that its moving surface center is constant also is one of starting point of the present invention.

For said method more clearly is described, describe a preferred embodiment of video intrusion detection method of the present invention below in detail.

Suppose that in a certain practical application scene, the pixel resolution of video is w * h, then set up have identical initial value with reference to matrix M ₁With the comparison and detection matrix M ₂The two-dimensional matrix that is w * h.In order to reduce the data volume of processing, in the present embodiment, make M ₁=M ₂=0, that is to say, under the original state, with reference to matrix M ₁With the comparison and detection matrix M ₂In all elements be 0.

One, sets up with reference to matrix

Suppose that in initialization cycle by calculating in one two field picture of back the poor of respective pixel point value in the value of each pixel and former frame image, the frame difference matrix that obtains is:

${\left[\begin{array}{cccccccccccccccccccc}0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& \·\·\·& 0\\ 0& & & & & & & & & & & & & & & & & & & 0\\ 0& & 0& & & & & & & & & & & & & & 0& & & 0\\ 0& & & & & 4& & & & & & & & & & & & & & 0\\ 0& & & & 2& 2& 3& & & & & & & & & 0& & & & 0\\ 0& & & 3& 1& 3& 6& 2& & & & & & & & & & & & 0\\ 0& & 1& 3& 2& 6& 5& 8& 2& & & & & & & & & & & 0\\ 0& & & & & & & & & & & & & & & & & & & 0\\ 0& & & & & & & & & & & & & 1& 6& 2& & & & 0\\ 0& & & & & & & & & & & & & 2& 4& 8& & & & 0\\ 0& & & & & & & & & & & & & 3& 2& 5& & & & 0\\ 0& & & & 0& & & & & & & & & & & & & & & 0\\ 0& & & & & & & & & & & & & & & & & & & 0\\ \·\·\·& & & & & & & & & & & & & & & & & & & \·\·\·\\ 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& \·\·\·& 0\end{array}\right]}_{(w\×h)}$

As implied above, in this frame difference matrix, except non-zero number, the element in this matrix is zero.By this frame difference matrix, can obtain two independently moving surfaces, moving surface A ₁With moving surface A ₂,

The position of getting first element place, the upper left corner is an origin, moving surface A ₁The coordinate of geometric center be: (6,6), moving surface A ₂The coordinate of geometric center is (8,15).Certainly, determine the center of moving surface according to the set center, a kind of preferred implementation wherein just, the method at other definite moving surface center also can adopt, and does not do qualification inferior.

Then, according to the coordinate of the geometric center of the above-mentioned moving surface that obtains, at the first statistical matrix M of correspondence ₁In, (6,6) point and (8,15) point are changed to 1.

Next, repeat to obtain frame difference matrix, obtain separation moving surface, obtain the pairing coordinate of geometric center of disengaging movement face and in the reference matrix M ₁In, the element at same coordinate place is changed to 1.

Above-mentioned steps constantly is repeated to carry out, until finishing an initialization cycle.As previously described, in actual applications, this initialization cycle should be a relatively long time, can be 10 minutes.

Suppose after initialization cycle is finished, to be confirmed as with reference to matrix:

$M_1={\left[\begin{array}{cccccccccccccc}& & & & & & & & & & 1& 1& & \·\·\·\\ & & & & & & & & & 1& 1& 1& 1& \\ & 1& 1& 1& & & & & & 1& 1& 1& 1& \\ & 1& 1& 1& 1& & & & & 1& 1& 1& 1& \\ & 1& 1& 1& 1& 1& & & & & & & & \\ & 1& 1& 1& & & & & & & 1& & & \\ & & & & & & & & 1& 1& 1& 1& & \\ & & & & & & & & 1& 1& 1& 1& 1& \\ \·\·\·& & & & & & & & & 1& 1& 1& & \·\·\·\\ & & & & & & & & & & & & & \·\·\·\end{array}\right]}_{w\×h}$

Wherein, there is not the element of mark part to be 0.

Two, set up the comparison and detection matrix

With with setting up, set up the comparison and detection matrix with reference to the identical method of matrix.Difference is, this comparison and detection matrix to set up period ratio shorter, generally set up, such as 2 frames by several frames.

Suppose certain Video Detection cycle then after, be confirmed as with reference to matrix:

$M_2={\left[\begin{array}{cccccccccccccc}& & & & & & & & & & 1& 1& & \·\·\·\\ & & & & & & & & & 1& 1& 1& 1& \\ & 1& & & & & & & & 1& 1& 1& 1& \\ & 1& 1& 1& & & & & & & 1& 1& & \\ & 1& 1& 1& 1& & & & & & & & & \\ & & & & & & & & & & 1& & & \\ & & & & & & & & & 1& 1& 1& & \\ & & & & & & & & 1& 1& 1& 1& 1& \\ \·\·\·& & & & & & & & & 1& 1& 1& & \·\·\·\\ & & & & & & & & & & & & \·\·\·& \end{array}\right]}_{w\×h}$

Wherein, there is not the element of mark part to be 0.

Three, compare with reference to matrix and described comparison and detection matrix described, and, determine the invasion object according to comparative result.

Below in conjunction with top M ₁Matrix and M ₂Matrix specifies how to have judged whether the invasion object: suppose that preassigned variance rate threshold value is 20%, and the computing formula of variance rate is defined as:

From M through mark ₁The distribution of matrix element comprises three abnormal areas, S as can be seen ₁, S ₂And S ₃, it is at M ₂Zone corresponding in the matrix is respectively: S ₁₁, S ₂₁, S ₃₁, make S ₁With S ₁₁Between variance rate by D ₁Represent S ₂With S ₂₁Between variance rate by D ₂Represent; S ₃With S ₃₁Between variance rate by D ₂Represent; And, P ₁Expression S ₁With S ₁₁Between, the number of the different elements of coordinate, P ₂Expression S ₂With S ₂₁Between, the number of the different elements of coordinate, P ₃Expression S ₃With S ₃₁Between, the number of the different elements of coordinate, then:

By aforementioned calculation, can draw: at video S ₁In the zone, the invasion object has appearred, and at S ₂, S ₃In the zone, do not invade the appearance of object.

Afterwards, with M ₁Reproduction matrix is to M ₂In the matrix, make M then ₁Matrix is reset to 0.Advance then as arriving the next video intrusion detection cycle.

By above-mentioned analysis, as can be seen, the present invention makes full use of because natural cause causes the motion clocklike that produces, the principle that its moving surface center is constant, by setting up with reference to matrix and comparison and detection matrix, compare with reference to matrix and described comparison and detection matrix described then, and according to comparative result, determine the technological means of invasion object, overcome in the prior art, detect the defective of the low precision that is brought based on frame of video pixel point frame difference, and then eliminate the motion artifacts clocklike that those produce owing to natural cause automatically, and then improved the precision of intrusion detection.

With reference to Fig. 5, Fig. 5 is the structural representation of the video detecting device embodiment according to the present invention, and according to embodiments of the invention, the present invention also provides a kind of video invasion detecting device, and this video invasion detecting device comprises:

Preset parameter unit 501, be used to preset have identical initial value with reference to matrix and comparison and detection matrix, and described size with reference to matrix and described comparison and detection matrix is corresponding with the resolution of described video; Preset initialization cycle and video intrusion detection cycle;

Set up unit 502 with reference to matrix, be used in described initialization cycle, set up described with reference to matrix;

The comparison and detection matrix is set up unit 503, is used for setting up described comparison and detection matrix in the described video intrusion detection cycle;

Invasion object determining unit 504 is used for comparing with reference to matrix and described comparison and detection matrix described, and according to comparative result, determines the invasion object.

With reference to Fig. 6, Fig. 6 is according to the present invention in the video detecting device, sets up the structural representation of unit, unit with reference to matrix, comprising:

The first frame difference matrix acquisition module 601 is used for calculating the poor of corresponding pixel points in the value of back one each pixel of two field picture and the former frame image, obtains frame difference matrix;

The first moving surface acquisition module 602 is used for obtaining the moving surface of a plurality of separation according to described frame difference matrix;

First is provided with module 603, is used to calculate the center of described each moving surface, and is provided with according to preset rules with reference to the element corresponding to described center position on the matrix described;

First judge module 604 is used to judge whether described initialization cycle is overtime, if not, then returns the described step of obtaining frame difference matrix.

With reference to Fig. 7, Fig. 7 is according to the present invention in the video detecting device, and the comparison and detection matrix is set up the structural representation of unit, unit, comprising:

The second frame difference matrix acquisition module 701 is used for calculating the poor of corresponding pixel points in the value of back one each pixel of two field picture and the former frame image, obtains frame difference matrix;

The second moving surface acquisition module 702 is used for obtaining the moving surface of a plurality of separation according to described frame difference matrix;

Second is provided with module 703, is used to calculate the center of described each moving surface, and the element corresponding to described center position on the described comparison and detection matrix is provided with according to preset rules;

Second judge module 704 is used to judge whether the described video intrusion detection cycle is overtime, if not, and the gating second frame difference matrix acquisition module then.

With reference to Fig. 8, Fig. 8 in the video detecting device, invades the structural representation of object determining unit according to the present invention, comprising:

Variance rate threshold value preset module 801 is used to preset the variance rate threshold value;

Abnormal area determination module 802 is used for determining abnormal area according to described distribution with reference to the matrix element;

Module 803 is searched in the corresponding region, is used at described comparison and detection matrix, searches and the corresponding zone of described abnormal area;

Variance rate computing module 804 is used to add up between described abnormal area and the described corresponding zone difference of element coordinate, and calculated difference rate;

The 3rd judge module 805 is used to judge that whether described variance rate is greater than described variance rate threshold value, if then this abnormal area has the invasion object; If not, then this abnormal area is not invaded object.

Preferably, described with reference to the matrix initial value be 0, described comparison and detection matrix initial value is 0.

Preferably, in the described indexing unit, describedly be specially with reference to carrying out mark corresponding to the element of described center position according to preset rules on the matrix: describedly be set to 1 with reference to the element corresponding to described center position on the matrix with described.

Preferably, above-mentioned center is the geometric center of moving surface.

According to embodiments of the invention, the present invention also provides a kind of chip that is used for the video intrusion detection, and described chip comprises above-mentioned video invasion detecting device.The structure of video invasion detecting device wherein is with above-described identical.Do not repeat them here.

To sum up, in the present invention:

The present invention makes full use of because natural cause causes the motion clocklike that produces, the principle that its moving surface center is constant, by setting up with reference to matrix and comparison and detection matrix, compare with reference to matrix and described comparison and detection matrix described then, and according to comparative result, determine the technological means of invasion object, overcome in the prior art, detect the defective of the low precision that is brought based on frame of video pixel point frame difference, and then eliminate the motion artifacts clocklike that those produce owing to natural cause automatically, and then improved the precision of intrusion detection.

Each embodiment in this instructions all adopts the mode of going forward one by one to describe, and what each embodiment stressed all is and the difference of other embodiment that identical similar part is mutually referring to getting final product between each embodiment.For system embodiment, because it is similar substantially to method embodiment, so description is fairly simple, relevant part gets final product referring to the part explanation of method embodiment.

More than to a kind of intrusion detection method provided by the present invention, pick-up unit, chip based on video, be described in detail, used specific case herein principle of the present invention and embodiment are set forth, the explanation of above embodiment just is used for helping to understand method of the present invention and core concept thereof; Simultaneously, for one of ordinary skill in the art, according to thought of the present invention, the part that all can change in specific embodiments and applications, in sum, this description should not be construed as limitation of the present invention.

Claims (19)

1, a kind of video intrusion detection method is characterized in that, preset have identical initial value with reference to matrix and comparison and detection matrix, and described size with reference to matrix and described comparison and detection matrix is corresponding with the resolution of described video; Preset initialization cycle and video intrusion detection cycle; Described method comprises the steps:

In described initialization cycle, set up described with reference to matrix;

In the described video intrusion detection cycle, set up described comparison and detection matrix;

Compare with reference to matrix and described comparison and detection matrix described, and, determine the invasion object according to comparative result.

2, method according to claim 1 is characterized in that, in described initialization cycle, sets up and describedly further comprises the steps: with reference to matrix

Calculate value poor of corresponding pixel points in the value of each pixel in one two field picture of back and the former frame image, obtain frame difference matrix;

According to described frame difference matrix, obtain the moving surface of a plurality of separation;

Calculate the center of described each moving surface, and be provided with according to preset rules with reference to the element corresponding to described center position on the matrix described;

Do not have under the overtime situation at described initialization cycle, return and carry out the described step of obtaining frame difference matrix.

3, method according to claim 2 is characterized in that, described initialization cycle is greater than the described video intrusion detection cycle.

4, method according to claim 3 is characterized in that, in the described video intrusion detection cycle, sets up described comparison and detection matrix, further comprises the steps:

Calculate the poor of corresponding pixel points in the value of each pixel in one two field picture of back and the former frame image, obtain frame difference matrix;

According to described frame difference matrix, obtain the moving surface of a plurality of separation;

Calculate the center of described each moving surface, and the element corresponding to described center position on the described comparison and detection matrix is provided with according to preset rules;

Judge whether the described video intrusion detection cycle is overtime, if not, then return the described step of obtaining frame difference matrix.

5, detection method according to claim 4 is characterized in that, compares with reference to matrix and described comparison and detection matrix described, and according to comparative result, determines the invasion object, further comprises the steps:

Preset the variance rate threshold value;

According to described distribution, determine abnormal area with reference to element in the matrix;

In described comparison and detection matrix, search and the corresponding zone of described abnormal area;

Add up between described abnormal area and the described corresponding zone difference of element coordinate, and calculated difference rate;

Judge that whether described variance rate is greater than described variance rate threshold value, if then this abnormal area has the invasion object; If not, then this abnormal area is not invaded object.

6, detection method according to claim 5 is characterized in that, described R-matrix initial value is 0, described comparison and detection matrix initial value is 0.

7, detection method according to claim 6, it is characterized in that, the center of described each moving surface of described calculating, and with described with reference to being provided with further according to preset rules corresponding to the element of described center position on the matrix be: the element corresponding to described center position on the described R-matrix is set to 1.

8, detection method according to claim 7 is characterized in that, described center is the geometric center of moving surface.

9, a kind of video invasion detecting device is characterized in that, comprising:

The preset parameter unit, be used to preset have identical initial value with reference to matrix and comparison and detection matrix, and described size with reference to matrix and described comparison and detection matrix is corresponding with the resolution of described video; Preset initialization cycle and video intrusion detection cycle;

Set up the unit with reference to matrix, be used in described initialization cycle, set up described with reference to matrix;

The comparison and detection matrix is set up the unit, is used for setting up described comparison and detection matrix in the described video intrusion detection cycle;

Invasion object determining unit is used for comparing with reference to matrix and described comparison and detection matrix described, and according to comparative result, determines the invasion object.

10, device according to claim 9 is characterized in that, describedly sets up the unit with reference to matrix and further comprises:

The first frame difference matrix acquisition module is used for calculating the poor of corresponding pixel points in the value of back one each pixel of two field picture and the former frame image, obtains frame difference matrix;

The first moving surface acquisition module is used for obtaining the moving surface of a plurality of separation according to described frame difference matrix;

First is provided with module, is used to calculate the center of described each moving surface, and is provided with according to preset rules with reference to the element corresponding to described center position on the matrix described;

First judge module is used to judge whether described initialization cycle is overtime, if not, and the gating first frame difference matrix acquisition module then.

11, device according to claim 10 is characterized in that, described comparison and detection matrix is set up the unit and further comprised:

The second frame difference matrix acquisition module is used for calculating the poor of corresponding pixel points in the value of back one each pixel of two field picture and the former frame image, obtains frame difference matrix;

The second moving surface acquisition module is used for obtaining the moving surface of a plurality of separation according to described frame difference matrix;

Second is provided with module, is used to calculate the center of described each moving surface, and the element corresponding to described center position on the described comparison and detection matrix is provided with according to preset rules;

Second judge module is used to judge whether the described video intrusion detection cycle is overtime, if not, and the gating second frame difference matrix acquisition module then.

12, device according to claim 11 is characterized in that, described invasion object determining unit further comprises:

Variance rate threshold value preset module is used to preset the variance rate threshold value;

The abnormal area determination module is used for determining abnormal area according to described distribution with reference to the matrix element;

Module is searched in the corresponding region, is used at described comparison and detection matrix, searches and the corresponding zone of described abnormal area;

The variance rate computing module is used to add up between described abnormal area and the described corresponding zone difference of element coordinate, and calculated difference rate;

The 3rd judge module is used to judge that whether described variance rate is greater than described variance rate threshold value, if then this abnormal area has the invasion object; If not, then this abnormal area is not invaded object.

13, pick-up unit according to claim 12 is characterized in that, described R-matrix initial value is 0, described comparison and detection matrix initial value is 0.

14, pick-up unit according to claim 13, it is characterized in that, the center of described each moving surface of described calculating, and be specially according to preset rules setting with reference to the element corresponding to described center position on the matrix described: the element corresponding to described center position on the described R-matrix is set to 1.

15, pick-up unit according to claim 14 is characterized in that, described center is the geometric center of moving surface.

16, a kind of chip that is used for the video intrusion detection, described chip comprises the video invasion detecting device, it is characterized in that, described video invasion detecting device comprises:

The preset parameter unit, be used to preset have identical initial value with reference to matrix and comparison and detection matrix, and described size with reference to matrix and described comparison and detection matrix is corresponding with the resolution of described video; Preset initialization cycle and video intrusion detection cycle;

Set up the unit with reference to matrix, be used in described initialization cycle, set up described with reference to matrix;

The comparison and detection matrix is set up the unit, is used for setting up described comparison and detection matrix in the described video intrusion detection cycle;

Invasion object determining unit is used for comparing with reference to matrix and described comparison and detection matrix described, and according to comparative result, determines the invasion object.

17, chip according to claim 16 is characterized in that, describedly sets up the unit with reference to matrix and further comprises:

The first frame difference matrix acquisition module is used for calculating the poor of corresponding pixel points in the value of back one each pixel of two field picture and the former frame image, obtains frame difference matrix;

The first moving surface acquisition module is used for obtaining the moving surface of a plurality of separation according to described frame difference matrix;

First is provided with module, is used to calculate the center of described each moving surface, and is provided with according to preset rules with reference to the element corresponding to described center position on the matrix described;

First judge module is used to judge whether described initialization cycle is overtime, if not, then returns the described step of obtaining frame difference matrix.

18, chip according to claim 17 is characterized in that, described comparison and detection matrix is set up the unit and further comprised:

The second frame difference matrix acquisition module is used for calculating the poor of corresponding pixel points in the value of back one each pixel of two field picture and the former frame image, obtains frame difference matrix;

The second moving surface acquisition module is used for obtaining the moving surface of a plurality of separation according to described frame difference matrix;

Second is provided with module, is used to calculate the center of described each moving surface, and the element corresponding to described center position on the described comparison and detection matrix is provided with according to preset rules;

Second judge module is used to judge whether the described video intrusion detection cycle is overtime, if not, and the gating second frame difference matrix acquisition module then.

19, chip according to claim 18 is characterized in that, described invasion object determining unit further comprises:

Variance rate threshold value preset module is used to preset the variance rate threshold value;

The abnormal area determination module is used for determining abnormal area according to described distribution with reference to the matrix element;

Module is searched in the corresponding region, is used at described comparison and detection matrix, searches and the corresponding zone of described abnormal area;

The variance rate computing module is used to add up between described abnormal area and the described corresponding zone difference of element coordinate, and calculated difference rate;

The 3rd judge module is used to judge that whether described variance rate is greater than described variance rate threshold value, if then this abnormal area has the invasion object; If not, then this abnormal area is not invaded object.

Images