CN103501428A - Multimedia monitoring method and multimedia monitoring device - Google Patents
Multimedia monitoring method and multimedia monitoring device Download PDFInfo
- Publication number
- CN103501428A CN103501428A CN201310499236.XA CN201310499236A CN103501428A CN 103501428 A CN103501428 A CN 103501428A CN 201310499236 A CN201310499236 A CN 201310499236A CN 103501428 A CN103501428 A CN 103501428A
- Authority
- CN
- China
- Prior art keywords
- macro block
- field picture
- vector
- multimedia
- sub
- 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
Links
Images
Landscapes
- Image Analysis (AREA)
Abstract
The invention provides a multimedia monitoring method, which includes the following steps: a frame image is acquired and decoded into a needed format; the frame image is segmented into a plurality of macroblocks; each macroblock is segmented into a plurality of sub-macroblocks; the characteristic value of each sub-macroblock is extracted in order to obtain the sample matrixes of the corresponding macroblocks; according to the sample matrixes, the macroblock characteristic vectors of the corresponding macroblocks are obtained; and according to all the macroblock characteristic vectors of the frame image, the abnormality of the frame image is detected. The monitoring method can accurately locate the specific positions of differences in the frame image. The invention also provides a multimedia monitoring device.
Description
Technical field
The present invention relates to a kind of multimedia monitoring method and device, particularly, method and the device of the image characteristics extraction based on vector, image comparison and abnormality detection technology.
Background technology
The extraction of characteristics of image, image ratio are technology commonly used in multimedia monitoring to reaching the image abnormity detection.Existing image ratio pair, can not navigate to the particular location of difference, can't find out the particular location of image difference.
In multimedia monitoring, image characteristics extraction is core technology.The method that three kinds of image characteristics extractions are arranged at present: one, the method based on statistics with histogram; Its two, the extracting method based on the Zernike square; Its three, the Sift Feature Correspondence Algorithm.There is following defect in these three kinds of image characteristic extracting methods: if computational accuracy is inadequate, can cause so feature extraction inaccurate; If the raising computation complexity, can make the real-time variation again.
In addition, existing multimedia monitoring technology is not strong to the adaptability of aspect ratio, can't meet the requirement of yardstick and swing consistency.
Summary of the invention
For solving the problems of the technologies described above, the invention provides a kind of multimedia monitoring method and device, it can navigate to the particular location of difference exactly.
The technical scheme adopted solved the problems of the technologies described above is to provide a kind of multimedia monitoring method, comprising:
Obtain two field picture, and described two field picture is decoded into to required form;
Described two field picture is divided into to a plurality of macro blocks;
Each described macroblock partition is become to a plurality of sub-macro blocks;
Extract the characteristic value of each described sub-macro block, to obtain the sampling matrix of corresponding macro block;
Obtain the macro block characteristics vector of corresponding macro block according to described sampling matrix;
Carry out the two field picture abnormality detection according to all described macro block characteristics vector of described two field picture.
Wherein, in the described step that described two field picture is divided into to a plurality of macro blocks, according to the Aspect Ratio of described two field picture, be evenly divided into a plurality of macro blocks.
Wherein, during described macroblock partition becomes the step of a plurality of sub-macro blocks by each, each described macroblock partition is become to four sub-macro blocks described.
Wherein, in the characteristic value step of described each described sub-macro block of extraction, in the same position of described sub-macro block, extract characteristic value.
Wherein, in the characteristic value step of described each described sub-macro block of extraction, described characteristic value is brightness value.
Wherein, in the described macro block characteristics vector that obtains corresponding macro block according to described sampling matrix, by obtain described macro block characteristics vector with down conversion, described macro block characteristics vector comprises brightness variable quantity λ and brightness change direction θ;
Wherein: x and y are obtained by following formula,
x＝|a
0.0-a
0.1|+…+|a
n.m-a
n.m+1|
y＝|a
0.0-a
1.0|+…+|a
n+1.m-a
n+1.m+1|
A
n.mcome from sampling matrix f
Wherein, carry out two field picture abnormality detection step according to described macro block characteristics vector and comprise described: the synchronous comparison step of the positioning step that two field picture is abnormal and/or multi-channel video.
Wherein, in abnormal positioning step, when meeting predefined difference condition, described macro block characteristics vector judges that this macro block occurs abnormal at two field picture.
Wherein, the synchronous comparison step of described multi-channel video comprises:
Obtain the set of the frame sequence characteristic vector of two-path video;
Obtain the step-length between the set of frame sequence characteristic vector of described two-path video;
Compare successively the described frame sequence characteristic vector of described two-path video, if the difference of described frame sequence characteristic vector reaches predetermined threshold, the image difference counting is added to 1.
Wherein, the synchronous comparison step of described multi-channel video also comprises:
Set alarm threshold value, if described image difference counting surpasses described alarm threshold value, give the alarm.
The present invention also provides a kind of multimedia monitor device, comprising:
Image extraction unit, it is for obtaining the two field picture of monitor video;
The image decoding unit, it is for being decoded into required form by described two field picture;
The characteristic value extraction unit, it is for described two field picture is divided into to a plurality of macro blocks, then each described macroblock partition is become to sub-macro block, extracts the characteristic value of each described sub-macro block, to obtain the sampling matrix of corresponding macro block;
The macro block characteristics vector calculation unit, it is for obtaining the macro block characteristics vector of corresponding macro block according to described sampling matrix;
The image abnormity detecting unit, it is for carrying out the two field picture abnormality detection according to described macro block characteristics vector.
Wherein, described characteristic value extraction unit obtains a plurality of macro blocks according to the Aspect Ratio even partition of described two field picture.
Wherein, described characteristic value extraction unit becomes four sub-macro blocks by each described macroblock partition.
Wherein, described characteristic value extraction unit extracts characteristic value in the same position of described sub-macro block.
Wherein, the described characteristic value that described characteristic value extraction unit extracts is brightness value.
Wherein, described macro block characteristics vector calculation unit is by obtaining described macro block characteristics vector with down conversion, and described macro block characteristics vector comprises brightness variable quantity λ and brightness change direction θ;
Wherein: x and y are obtained by following formula,
x＝|a
0.0-a
0.1|+…+|a
n.m-a
n.m+1|
y＝|a
0.0-a
1.0|+…+|a
n+1.m-a
n+1.m+1|
A
n.mcome from sampling matrix f:
Wherein, described image abnormity detecting unit comprises:
Two field picture difference locator unit, it judge that when described macro block characteristics vector meets predefined difference condition this macro block occurs extremely for the particular location of locating frame image abnormity;
The synchronous comparer unit of multi-channel video, it is for the comparison of multiple paths of video images.
Wherein, the synchronous comparer unit of described multi-channel video comprises:
Frame sequence characteristic vector acquisition module, it is for the set of the frame sequence characteristic vector of obtaining the described video of two-way;
The step size computation module, it is for the step-length between the set of the frame sequence characteristic vector of obtaining described two-path video;
Frame sequence characteristic vector comparing module, it is for comparing successively described frame sequence characteristic vector;
Counting module, it reaches predetermined threshold for the difference in described frame sequence characteristic vector, the image difference counting is added to 1.
Wherein, described image difference judging unit also comprises:
The warning subelement, it makes to give the alarm at described image difference counting, surpassing alarm threshold value.
The present invention has following beneficial effect:
Multimedia monitoring method provided by the invention is when extracting characteristics of image, two field picture is divided into to a plurality of macro blocks, again macroblock partition is become to a plurality of sub-macro blocks, extract the characteristic value of every sub-macro block to obtain the sampling matrix of corresponding macro block, obtain the macro block characteristics vector by the sampling matrix, carry out the two field picture Difference test according to the macro block characteristics vector again, the method that this piecemeal extracts characteristics of image is the particular location of difference in the locating frame image accurately.
As an advantage of the present invention, the characteristic value of extraction is brightness value, based on brightness value, extracts feature, therefore, can adapt to the two field picture of different proportion, has improved the adaptability of two field picture ratio; And, there is yardstick consistency and swing consistency, and the advantage of occupied bandwidth less (video of HD only has 144B/frame).
As another advantage of the present invention, the algorithm that is obtained the macro block characteristics vector by the sampling matrix of macro block has moderate complexity, when guaranteeing precision, there is again good real-time, when dominant frequency is 1GHz, multimedia monitoring method of the present invention can be extracted 16 tunnels simultaneously, and existingly only can extract 4 tunnels.
Multimedia monitor device provided by the invention is divided into a plurality of macro blocks by the characteristic value extraction unit by two field picture, again macroblock partition is become to a plurality of sub-macro blocks, extract the characteristic value of every sub-macro block to obtain the sampling matrix of corresponding macro block, obtain the macro block characteristics vector by the sampling matrix, carry out the two field picture Difference test according to the macro block characteristics vector again, this multimedia monitor device is the particular location of difference in the locating frame image accurately.
The accompanying drawing explanation
The flow chart that Fig. 1 is embodiment of the present invention multimedia monitoring method;
Fig. 2 is that the embodiment of the present invention is divided macro block and sub-macro block and sampling example;
The diagram that Fig. 3 is embodiment of the present invention macro block characteristics vector means mode;
The characteristic vector diagram that Fig. 4 is black of the embodiment of the present invention;
The flow chart of the synchronous comparison that Fig. 5 is embodiment of the present invention multi-channel video;
Fig. 6 is the synchronous schematic diagram of embodiment of the present invention frame sequence;
The theory diagram that Fig. 7 is embodiment of the present invention multimedia monitor device;
The theory diagram of the synchronous comparer unit that Fig. 8 is embodiment of the present invention multi-channel video.
Embodiment
For making those skilled in the art understand better technical scheme of the present invention, below in conjunction with accompanying drawing, multimedia monitoring method provided by the invention and device are described in detail.
The present embodiment multimedia monitoring method and device are based on method and the device of image characteristics extraction, image comparison and the abnormality detection technology of vector.As shown in Figure 1, the present embodiment multimedia monitoring method comprises the following steps:
Step S11, obtain two field picture, and two field picture is decoded into to required form.
The multimedia monitoring image is comprised of a series of continuous two field pictures, and therefore, follow-up image characteristics extraction, image comparison and abnormality detection are all for two field picture.Obtain two field picture, and two field picture is decoded into to required form, as yuv format or rgb format.
Step S12, be divided into a plurality of macro blocks by two field picture.
Step S12 is for pressing the step of territorial sampling grouping.Two field picture is pressed to the area dividing grouping, be divided into a plurality of macro blocks.Because the Aspect Ratio of the two field picture obtained is generally 16:9 or 4:3, preferably according to the Aspect Ratio of two field picture, be evenly divided into a plurality of macro blocks.As shown in Figure 2, in figure, each grid means a macro block.The two field picture that is 16:9 by Aspect Ratio is divided into 16 * 9 macro blocks, and the two field picture that is 4:3 by Aspect Ratio is divided into 12 * 9 macro blocks.
Step S13, become a plurality of sub-macro blocks by each macroblock partition.
Step S13 is also for pressing the step of territorial sampling grouping.Each macro block further is divided into to a plurality of sub-macro blocks.As shown in Figure 2, each macro block further is divided into to 4 sub-macro blocks.Certainly, also each macroblock partition can be become to 9 or more sub-macro block.
Step S14, extract the characteristic value of every sub-macro block, to obtain the sampling matrix of corresponding macro block.
Step S14 is the sampling step.Extract the characteristic value Y of every sub-macro block, preferably in the same position of every sub-macro block, extract characteristic value Y, to obtain the sampling matrix of corresponding macro block.
In the present embodiment, the characteristic value Y of extraction is brightness value, based on brightness value, extracts feature, so not only can adapt to the two field picture of different proportion, has improved the adaptability of two field picture ratio; And, there is yardstick consistency and swing consistency, and the advantage of occupied bandwidth less (video of HD only has 144B/frame).
Step S15, obtain the macro block characteristics vector of corresponding macro block according to the sampling matrix.
The step that step S15 is the computing macro block characteristic vector.Particularly, the sampling matrix passes through to obtain the macro block characteristics vector with down conversion,
Wherein, λ means the brightness variable quantity, and θ means the brightness change direction;
Variable x and y are obtained by following formula:
x＝|a
0.0-a
0.1|+|a
1.0-a
1.1|
y＝|a
0.0-a
1.0|+|a
0.1-a
1.1|
Wherein, the diagramatic way of macro block characteristics vector as shown in Figure 3.This algorithm complex is moderate, when guaranteeing precision, has again good real-time, and when dominant frequency is 1GHz, multimedia monitoring method of the present invention can be extracted 16 tunnels simultaneously, and existingly only can extract 4 tunnels.
Step S16, carry out the two field picture abnormality detection according to all macro block characteristics vectors of two field picture.
The two field picture abnormality detection comprises positioning step that two field picture is abnormal and/or the synchronous comparison step of multi-channel video.
Judge that according to the macro block characteristics vector whether this macro block is abnormal, can judge that when the macro block characteristics vector meets predefined difference condition this macro block occurs abnormal.Due to a certain particular location in the corresponding two field picture of each macro block, therefore, by the macro block characteristics vector, just can locate exactly abnormal particular location.
For example, as shown in Figure 4, when detecting for black, the difference condition is set as to θ=45, λ≤16, if all macro block characteristics vectors of two field picture meet θ=45, λ≤16, be judged to be black.
The synchronous comparison of multi-channel video not only can synchronously be compared two-path video, also can to three roads or more multi-channel video synchronously compare.Only take the synchronous comparison of two-path video at this describes as example.Particularly, as shown in Figure 5, the synchronous comparison of multi-channel video comprises the following steps:
Step S61, obtain the frame sequence characteristic vector of two-path video.
Obtain the frame sequence characteristic vector of two-path video.As, the set of A road sequence of frames of video characteristic vector is A
1, A
2... A
n, the set of B road sequence of frames of video characteristic vector is B
1, B
2... B
m
Step S62, obtain the step-length between the frame sequence characteristic vector set of two-path video.
As shown in Figure 6, if A road sequence of frames of video characteristic vector A
nwith B road sequence of frames of video characteristic vector B
mequate, i.e. A
n=B
m, the step-length s=|n-m| between the set of the set of A road sequence of frames of video characteristic vector and B road sequence of frames of video characteristic vector so.
Step S63, compare described frame sequence characteristic vector successively, if the difference of described frame sequence characteristic vector reaches predetermined threshold, the image difference counting added to 1.
As shown in Figure 6, compare successively A road sequence of frames of video characteristic vector A
nwith B road sequence of frames of video characteristic vector B
n+sif, A road sequence of frames of video characteristic vector A
nwith B road sequence of frames of video characteristic vector B
n+sbetween difference reach predefined threshold value, image difference counting difcnt++ adds 1.
Preferably, the synchronous comparison of multi-channel video also comprises:
Step S64, set alarm threshold value, if described image difference counting surpasses described alarm threshold value, gives the alarm.
At first preset alarm threshold value, when the image difference counting of step S63 acquisition surpasses this alarm threshold value, give the alarm, be beneficial to the synchronous comparison of user's multi-channel video.
In benzene embodiment, the sampling matrix obtained by step S14 comprises 4 extraction values, and in fact, the sampling matrix can comprise more extraction values, obtains (21) matrix,
Accordingly, in step S15, variable x and y are obtained by following formula,
x＝|a
0.0-a
0.1|+…+|a
n.m-a
n.m+1|
y＝|a
0.0-a
1.0|+…+|a
n+1.m-a
n+1.m+1|
The multimedia monitoring method that the present embodiment provides is when extracting characteristics of image, two field picture is divided into to a plurality of macro blocks, again macroblock partition is become to a plurality of sub-macro blocks, extract the characteristic value of every sub-macro block to obtain the sampling matrix of corresponding macro block, obtain the macro block characteristics vector by the sampling matrix, carry out the two field picture Difference test according to the macro block characteristics vector again, the method that this piecemeal extracts characteristics of image is the particular location of difference in the locating frame image accurately.
And the characteristic value of extraction is brightness value, based on brightness value, extract feature, therefore, can adapt to the two field picture of different proportion, improved the adaptability of two field picture ratio; And, there is yardstick consistency and swing consistency, and the advantage of occupied bandwidth less (video of HD only has 144B/frame).
In addition, the algorithm that is obtained the macro block characteristics vector by the sampling matrix of macro block has moderate complexity, when guaranteeing precision, there is again good real-time, when dominant frequency is 1GHz, multimedia monitoring method of the present invention can be extracted 16 tunnels simultaneously, and existingly only can extract 4 tunnels.
The theory diagram that Fig. 6 is multimedia monitor device.As shown in Figure 7, multimedia monitor device comprises:
Image extraction unit 1, it is for obtaining the two field picture of monitor video.
The multimedia monitoring image is comprised of a series of continuous two field pictures, and therefore, follow-up image characteristics extraction, image comparison and abnormality detection are all for two field picture.
Characteristic value extraction unit 3, it is for two field picture being divided into to a plurality of macro blocks, then each macroblock partition is become to sub-macro block, extracts the characteristic value of every sub-macro block, to obtain the sampling matrix of corresponding macro block.
Characteristic value extraction unit 3 is pressed the territorial sampling grouping, is about to two field picture and presses the area dividing grouping, is divided into a plurality of macro blocks, then each macroblock partition is become to a plurality of sub-macro blocks, extracts the characteristic value of every sub-macro block, to obtain the sampling matrix of corresponding macro block.
Because the Aspect Ratio of the two field picture obtained is generally 16:9 or 4:3, preferably according to the Aspect Ratio of two field picture, be evenly divided into a plurality of macro blocks.As shown in Figure 2, in figure, each grid means a macro block.The two field picture that is 16:9 by Aspect Ratio is divided into 16 * 9 macro blocks, and the two field picture that is 4:3 by Aspect Ratio is divided into 12 * 9 macro blocks.Each macro block is divided into 4 sub-macro blocks.Certainly, also each macroblock partition can be become to 9 or more sub-macro block.
Characteristic value extraction unit 3 extracts the characteristic value Y of every sub-macro block, preferably in the same position of every sub-macro block, extracts characteristic value Y, to obtain the sampling matrix of corresponding macro block.
In the present embodiment, the characteristic value Y of extraction is brightness value, based on brightness value, extracts feature, so not only can adapt to the two field picture of different proportion, has improved the adaptability of two field picture ratio; And, there is yardstick consistency and swing consistency, and the advantage of occupied bandwidth less (video of HD only has 144B/frame).
Macro block characteristics vector calculation unit 4, it is for obtaining the macro block characteristics vector of corresponding macro block according to the sampling matrix.
Macro block characteristics vector calculation unit 4 computing macro block characteristic vectors, particularly, the sampling matrix passes through to obtain the macro block characteristics vector with down conversion,
Wherein, λ means the brightness variable quantity, and θ means the brightness change direction;
Variable x and y are obtained by following formula:
x＝|a
0.0-a
0.1|+|a
1.0-a
1.1|
y＝|a
0.0-a
1.0|+|a
0.1-a
1.1|
The moderate complexity of the algorithm of macro block characteristics vector calculation unit 4, when guaranteeing precision, have again good real-time, and when dominant frequency is 1GHz, multimedia monitoring method of the present invention can be extracted 16 tunnels simultaneously, and existingly only can extract 4 tunnels.
Image abnormity detecting unit 5, it is for carrying out the two field picture abnormality detection according to described macro block characteristics vector.
Image abnormity detecting unit 5 comprises:
Two field picture difference locator unit 51, it judge that when the macro block characteristics vector meets predefined difference condition this macro block occurs extremely for the particular location of locating frame image abnormity.
The operation principle of two field picture difference locator unit 51, referring to step S61 in the multimedia monitoring method, does not repeat them here.
The synchronous comparer unit 52 of multi-channel video, it is for the comparison of multiple paths of video images.
As shown in Figure 8, the synchronous comparer unit 52 of multi-channel video comprises:
Frame sequence characteristic vector acquisition module 521, it is for the set of the frame sequence characteristic vector of obtaining two-path video.
The set that frame sequence characteristic vector acquisition module 521 obtains A road sequence of frames of video characteristic vector is A
1, A
2... A
n, and the set of B road sequence of frames of video characteristic vector is B
1, B
2... B
m
Step size computation module 522, it is for the step-length between the set of the frame sequence characteristic vector of obtaining two-path video.
Step size computation module 522 is calculated the step-length s of the set of two-path video frame sequence characteristic vector.As shown in Figure 5, if A road sequence of frames of video characteristic vector A
nwith B road sequence of frames of video characteristic vector B
mequate, i.e. A
n=B
m, the step-length s=|n-m| between the set of the set of A road sequence of frames of video characteristic vector and B road sequence of frames of video characteristic vector so.
Frame sequence characteristic vector comparing module 523, it is for comparing successively the frame sequence characteristic vector.
Frame sequence characteristic vector comparing module 523 is compared A road sequence of frames of video characteristic vector A successively
nwith B road sequence of frames of video characteristic vector B
n+s.
Counting module 524, it reaches predetermined threshold for the difference in described frame sequence characteristic vector, the image difference counting is added to 1.
As A road sequence of frames of video characteristic vector A
nwith B road sequence of frames of video characteristic vector B
n+sbetween difference reach predefined threshold value, image difference counting difcnt++ adds 1.
Preferably, the synchronous comparer unit 52 of multi-channel video also comprises warning subelement 525, and it makes to give the alarm at the image difference counting, surpassing alarm threshold value.
The multimedia monitor device that the present embodiment provides is divided into a plurality of macro blocks by the characteristic value extraction unit by two field picture, again macroblock partition is become to a plurality of sub-macro blocks, extract the characteristic value of every sub-macro block to obtain the sampling matrix of corresponding macro block, obtain the macro block characteristics vector by the sampling matrix, carry out the two field picture Difference test according to the macro block characteristics vector again, this multimedia monitor device is the particular location of difference in the locating frame image accurately.
Be understandable that, above execution mode is only the illustrative embodiments adopted for principle of the present invention is described, yet the present invention is not limited thereto.For those skilled in the art, without departing from the spirit and substance in the present invention, can make various modification and improvement, these modification and improvement also are considered as protection scope of the present invention.
Claims (19)
1. a multimedia monitoring method, is characterized in that, comprising:
Obtain two field picture, and described two field picture is decoded into to required form;
Described two field picture is divided into to a plurality of macro blocks;
Each described macroblock partition is become to a plurality of sub-macro blocks;
Extract the characteristic value of each described sub-macro block, to obtain the sampling matrix of corresponding macro block;
Obtain the macro block characteristics vector of corresponding macro block according to described sampling matrix;
Carry out the two field picture abnormality detection according to all described macro block characteristics vector of described two field picture.
2. multimedia monitoring method according to claim 1, is characterized in that, in the described step that described two field picture is divided into to a plurality of macro blocks, according to the Aspect Ratio of described two field picture, is evenly divided into a plurality of macro blocks.
3. multimedia monitoring method according to claim 1, is characterized in that, described, during described macroblock partition becomes the step of a plurality of sub-macro blocks by each, each described macroblock partition become to four sub-macro blocks.
4. multimedia monitoring method according to claim 1, is characterized in that, in the characteristic value step of described each described sub-macro block of extraction, in the same position of described sub-macro block, extracts characteristic value.
5. multimedia monitoring method according to claim 1, is characterized in that, in the characteristic value step of described each described sub-macro block of extraction, described characteristic value is brightness value.
6. multimedia monitoring method according to claim 1, it is characterized in that, in the described macro block characteristics vector that obtains corresponding macro block according to described sampling matrix, by obtain described macro block characteristics vector with down conversion, described macro block characteristics vector comprises brightness variable quantity λ and brightness change direction θ;
Wherein: x and y are obtained by following formula,
x＝|a
0.0-a
0.1|+…+|a
n.m-a
n.m+1|
y＝|a
0.0-a
1.0|+…+|a
n+1.m-a
n+1.m+1|
A
n.mcome from sampling matrix f
。
7. multimedia monitoring method according to claim 1, is characterized in that, carries out two field picture abnormality detection step according to described macro block characteristics vector and comprise described: the synchronous comparison step of the positioning step that two field picture is abnormal and/or multi-channel video.
8. multimedia monitoring method according to claim 7, is characterized in that, at two field picture, in abnormal positioning step, judges that when described macro block characteristics vector meets predefined difference condition this macro block occurs abnormal.
9. multimedia monitoring method according to claim 7, is characterized in that, the synchronous comparison step of described multi-channel video comprises:
Obtain the set of the frame sequence characteristic vector of two-path video;
Obtain the step-length between the set of frame sequence characteristic vector of described two-path video;
Compare successively the described frame sequence characteristic vector of described two-path video, if the difference of described frame sequence characteristic vector reaches predetermined threshold, the image difference counting is added to 1.
10. multimedia monitoring method according to claim 9, is characterized in that, the synchronous comparison step of described multi-channel video also comprises:
Set alarm threshold value, if described image difference counting surpasses described alarm threshold value, give the alarm.
11. a multimedia monitor device, is characterized in that, comprising:
Image extraction unit, it is for obtaining the two field picture of monitor video;
The image decoding unit, it is for being decoded into required form by described two field picture;
The characteristic value extraction unit, it is for described two field picture is divided into to a plurality of macro blocks, then each described macroblock partition is become to sub-macro block, extracts the characteristic value of each described sub-macro block, to obtain the sampling matrix of corresponding macro block;
The macro block characteristics vector calculation unit, it is for obtaining the macro block characteristics vector of corresponding macro block according to described sampling matrix;
The image abnormity detecting unit, it is for carrying out the two field picture abnormality detection according to described macro block characteristics vector.
12. multimedia monitor device according to claim 11, is characterized in that, described characteristic value extraction unit obtains a plurality of macro blocks according to the Aspect Ratio even partition of described two field picture.
13. multimedia monitor device according to claim 11, is characterized in that, described characteristic value extraction unit becomes four sub-macro blocks by each described macroblock partition.
14. multimedia monitor device according to claim 11, is characterized in that, described characteristic value extraction unit extracts characteristic value in the same position of described sub-macro block.
15. multimedia monitor device according to claim 11, is characterized in that, the described characteristic value that described characteristic value extraction unit extracts is brightness value.
16. multimedia monitor device according to claim 11, is characterized in that, described macro block characteristics vector calculation unit is by obtaining described macro block characteristics vector with down conversion, and described macro block characteristics vector comprises brightness variable quantity λ and brightness change direction θ;
Wherein: x and y are obtained by following formula,
x＝|a
0.0-a
0.1|+…+|a
n.m-a
n.m+1|
y＝|a
0.0-a
1.0|+…+|a
n+1.m-a
n+1.m+1|
A
n.mcome from sampling matrix f:
17. multimedia monitor device according to claim 11, is characterized in that, described image abnormity detecting unit comprises:
Two field picture difference locator unit, it judge that when described macro block characteristics vector meets predefined difference condition this macro block occurs extremely for the particular location of locating frame image abnormity;
The synchronous comparer unit of multi-channel video, it is for the comparison of multiple paths of video images.
18. multimedia monitor device according to claim 17, is characterized in that, the synchronous comparer unit of described multi-channel video comprises:
Frame sequence characteristic vector acquisition module, it is for the set of the frame sequence characteristic vector of obtaining the described video of two-way;
The step size computation module, it is for the step-length between the set of the frame sequence characteristic vector of obtaining described two-path video;
Frame sequence characteristic vector comparing module, it is for comparing successively described frame sequence characteristic vector;
Counting module, it reaches predetermined threshold for the difference in described frame sequence characteristic vector, the image difference counting is added to 1.
19. multimedia monitor device according to claim 17, is characterized in that, described image difference judging unit also comprises:
The warning subelement, it makes to give the alarm at described image difference counting, surpassing alarm threshold value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310499236.XA CN103501428A (en) | 2013-10-22 | 2013-10-22 | Multimedia monitoring method and multimedia monitoring device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310499236.XA CN103501428A (en) | 2013-10-22 | 2013-10-22 | Multimedia monitoring method and multimedia monitoring device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103501428A true CN103501428A (en) | 2014-01-08 |
Family
ID=49866593
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310499236.XA Pending CN103501428A (en) | 2013-10-22 | 2013-10-22 | Multimedia monitoring method and multimedia monitoring device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103501428A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107610721A (en) * | 2017-10-13 | 2018-01-19 | 北京奇虎科技有限公司 | Multimedia messages detection method and device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101004791A (en) * | 2007-01-19 | 2007-07-25 | 赵力 | Method for recognizing facial expression based on 2D partial least square method |
| CN101600110A (en) * | 2009-07-14 | 2009-12-09 | 杭州华三通信技术有限公司 | Method for testing motion and device |
| CN101706944A (en) * | 2009-11-03 | 2010-05-12 | 上海大学 | Quantization table evaluation based method for detecting JPEG image tampering |
| CN102348128A (en) * | 2010-07-30 | 2012-02-08 | 株式会社日立制作所 | Surveillance camera system having camera malfunction detection function |
-
2013
- 2013-10-22 CN CN201310499236.XA patent/CN103501428A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101004791A (en) * | 2007-01-19 | 2007-07-25 | 赵力 | Method for recognizing facial expression based on 2D partial least square method |
| CN101600110A (en) * | 2009-07-14 | 2009-12-09 | 杭州华三通信技术有限公司 | Method for testing motion and device |
| CN101706944A (en) * | 2009-11-03 | 2010-05-12 | 上海大学 | Quantization table evaluation based method for detecting JPEG image tampering |
| CN102348128A (en) * | 2010-07-30 | 2012-02-08 | 株式会社日立制作所 | Surveillance camera system having camera malfunction detection function |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107610721A (en) * | 2017-10-13 | 2018-01-19 | 北京奇虎科技有限公司 | Multimedia messages detection method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102842128B (en) | Method for checking object and device thereof | |
| Wang et al. | Novel spatio-temporal structural information based video quality metric | |
| CN106507372A (en) | A kind of wireless network zone-skipping coverage detection method, device and communication system | |
| CN103744005A (en) | Discharging strength conversion method based on ultraviolet imaging method | |
| CN106126385A (en) | A kind of unit exception real-time detection method based on synchronous data flow compression | |
| CN102842203A (en) | Method for monitoring bridge fault on basis of video image | |
| CN111950510B (en) | High-voltage switch switching indication board image recognition method | |
| CN113873229B (en) | Image dead pixel detection method, system and device | |
| CN100469153C (en) | Testing method of switching video scenes | |
| CN108174198B (en) | Video image quality diagnosis analysis detection device and application system | |
| CN208273026U (en) | A kind of environment measuring network system based on wireless sensor | |
| CN105743223A (en) | GOOSE linkage-based video inspection method for secondary equipment prefabricated cabin | |
| CN103501428A (en) | Multimedia monitoring method and multimedia monitoring device | |
| CN104424633A (en) | Video contrast abnormity detection method and device | |
| CN102595186A (en) | Method for identifying original video frame rate | |
| KR20120116699A (en) | Moving object detecting apparatus and method using clustering | |
| CN112418226B (en) | Method and device for identifying opening and closing states of fisheyes | |
| CN104808039A (en) | Automatic correction system and automatic correction method of electronic equipment | |
| CN103347141B (en) | A kind of CMYK four-color process aberration localization method | |
| KR102177494B1 (en) | method of identifying abnormal-motion objects in compressed video by use of trajectory and pattern of motion vectors | |
| CN105486930B (en) | With silicon chain direct current system bus to the detection device and method of ground resistance | |
| CN103983239A (en) | Distance measuring method based on lane equal-width line | |
| CN103607592B (en) | Video tampering detection method and system | |
| CN114460360B (en) | Detection method, system and device based on ammeter measurement current time integral | |
| CN114666525B (en) | Audio and video switching verification system based on ASIC structure |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| AD01 | Patent right deemed abandoned |
Effective date of abandoning: 20170908 |
|
| AD01 | Patent right deemed abandoned |