CN103561232A - Method and device for identifying video monitoring scenes - Google Patents
Method and device for identifying video monitoring scenes Download PDFInfo
- Publication number
- CN103561232A CN103561232A CN201310491808.XA CN201310491808A CN103561232A CN 103561232 A CN103561232 A CN 103561232A CN 201310491808 A CN201310491808 A CN 201310491808A CN 103561232 A CN103561232 A CN 103561232A
- Authority
- CN
- China
- Prior art keywords
- scene
- video
- frame
- video monitoring
- image
- 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
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
The invention discloses a method and device for identifying video monitoring scenes. The method for identifying the video monitoring scenes comprises the steps of (1) obtaining a front video image frame and a current video image frame obtained through monitoring of a front-end video monitoring device; (2) carrying out subtraction between the front video image frame and the obtained current video image frame to obtain an image residual frame; (3) determining brightness values of pixels of the obtained image residual frame; (4) determining the value of the ratio of the number of the pixels with the nonzero brightness values to the number of all the pixels of the image residual frame according to the determined brightness value of each pixel; (5) determining that the current monitoring scene where the front-end video monitoring device is located is a moving scene when the determined value of the ratio is larger than a set ratio threshold value; (6) determining that the current monitoring scene where the front-end video monitoring device is located is a static scene when the determined value of the ratio is smaller than or equal to the set ratio threshold value. The method and device for identifying the video monitoring scenes can achieve identification of the different monitoring scenes.
Description
Technical field
The present invention relates to technical field of video monitoring, especially relate to a kind of video monitoring scene method of discrimination and device.
Background technology
At present, be accompanied by the Large scale construction of the maturation of security protection industry and safe city, Safe Campus, real-time remote monitoring more and more obtains people's attention, and video monitoring has obtained application more and more widely.Video monitoring also from wired develop into wireless, to meet ever-increasing mobility and convenience requirement.But because the upstream bandwidth of the 3G wireless networks such as TD-SCDMA is limited, make the transmission quality of video monitoring image not good, greatly affected the effect of video monitoring.
When increasing wireless network transmissions bandwidth, also need the Video coding mode of information source end to be optimized, to be lifted at the transmission of video images quality under prior wireless network situation.But generally the application scenarios of video monitoring is a lot, as daytime, night, the monitoring scene such as fixing, mobile, the video image characteristic that different monitoring scene lower front end camera head gathers is not identical, not identical to the performance requirement of encoder yet, therefore, need to identify the state of the application scenarios of current monitoring.
Summary of the invention
The embodiment of the present invention provides a kind of video monitoring scene method of discrimination and device thereof, to realize the identification of the application scenarios (moving scene/static scene) to current monitoring.
The technical scheme that the embodiment of the present invention proposes is as follows:
A method of discrimination, comprises and obtains last video frame image and the current video image frame that front end video monitoring apparatus monitors; The current video image frame of acquisition and last video frame image are subtracted each other and obtain Image Residual frame; Determine the brightness value of each pixel in the Image Residual frame obtaining; According to the brightness value of each pixel of determining, determine the ratio value of the number of all pixels in the pixel number of non-zero brightness value and described residual frame; If definite ratio value is greater than the proportion threshold value of setting, differentiating the current residing monitoring scene of front end video monitoring apparatus is moving scene; Otherwise differentiating the current residing monitoring scene of front end video monitoring apparatus is static scene.
A discriminating gear, comprises that picture frame obtains unit, the last video frame image and the current video image frame that for obtaining front end video monitoring apparatus, monitor; Residual frame obtains unit, for picture frame being obtained to current video image frame that unit obtains and last video frame image, subtracts each other and obtains Image Residual frame; Brightness value determining unit, for determining the brightness value of each pixel of Image Residual frame that residual frame acquisition unit obtains; Ratio value determining unit, for the brightness value of each pixel of determining according to brightness value determining unit, determines the ratio value of the number of all pixels in the pixel number of non-zero brightness value and described residual frame; Scene judgement unit, for when the definite ratio value of ratio value determining unit is greater than the proportion threshold value of setting, differentiating the current residing monitoring scene of front end video monitoring apparatus is moving scene; Otherwise differentiating the current residing monitoring scene of front end video monitoring apparatus is static scene.
The embodiment of the present invention, by the identifying schemes of utilization campaign/static monitoring scene, has promoted the video monitoring image quality under various monitoring scenes.
Accompanying drawing explanation
Fig. 1 is the composition structural representation of video monitoring system;
Fig. 2 is video coding process schematic diagram in video monitoring system;
Fig. 3 is Video coding framework schematic diagram detailed in video monitoring system;
Fig. 4 is the process chart that the differentiation video monitoring scene of embodiment of the present invention proposition is motion/static scene;
Fig. 5 is the composition structural representation of differentiation video monitoring scene that the embodiment of the present invention the proposes processing unit that is motion/static scene;
Fig. 6 is that the differentiation video monitoring scene that the embodiment of the present invention proposes is the process chart of scene at daytime/night;
Fig. 7 is that differentiation video monitoring scene that the embodiment of the present invention proposes is the composition structural representation of processing unit of scene at daytime/night;
Fig. 8 is that the differentiation video monitoring scene that the embodiment of the present invention proposes is the process chart of indoor/outdoor scene;
Fig. 9 is that the differentiation video monitoring scene that the embodiment of the present invention proposes is the composition structural representation of the processing unit of indoor/outdoor scene;
Figure 10 is that the first of embodiment of the present invention proposition is for the process chart of the video monitoring image coding method of moving scene;
Figure 11 is the composition structural representation of the first that the embodiment of the present invention proposes video monitoring image code processing apparatus that is moving scene for video monitoring scene;
Figure 12 is that the second of embodiment of the present invention proposition is for the enforcement schematic diagram of the video monitoring image coding method of moving scene;
Figure 13 is that the second of embodiment of the present invention proposition is for the process chart of the video monitoring image coding method of moving scene;
Figure 14 is the composition structural representation of the second that the embodiment of the present invention proposes video monitoring image code processing apparatus that is moving scene for video monitoring scene;
Figure 15 is the process chart of the video monitoring image coding method for static scene of embodiment of the present invention proposition;
Figure 16 is the composition structural representation of the video monitoring image code processing apparatus that is static scene for video monitoring scene of embodiment of the present invention proposition;
Figure 17 is the schematic diagram that the embodiment of the present invention is optimized and revised null value area quantization;
Figure 18 is code optimization schematic flow sheet under scene at night;
What Figure 19 was that the embodiment of the present invention proposes is the composition structural representation of video monitoring image code processing apparatus of scene at night on daytime for video monitoring scene;
Figure 20 is the composition structural representation of the video monitoring image code processing apparatus that is indoor scene for video monitoring scene of embodiment of the present invention proposition;
Figure 21 is the process chart of the video monitoring coding method of embodiment of the present invention proposition;
Figure 22 is the composition structural representation of the video monitoring code processing apparatus of embodiment of the present invention proposition.
Embodiment
For existing video monitoring scheme, all do not consider the different scene features of monitoring application, for different monitoring scenes, do not take different code optimization strategies, therefore can not be applicable to multiple monitoring scene.In addition, the video monitoring particularly kind of the front end watch-dog of wireless video monitoring is more, the disposal ability of a lot of front end watch-dogs a little less than, and adopt existing based on as H.264 the encoding scheme of standard is too complicated, on these equipment, be difficult to carry out real-time coding, or cannot carry out multiplex coding, thereby affected carrying out of Video Supervision Technique.Problem for prior art, this programme embodiment has proposed a set of code optimization scheme for video surveillance applications, video surveillance applications scene based on different is carried out code optimization targetedly, to be fully lifted at the video quality under each monitoring scene, provide more good user to experience, and and then reduce the video encoding complexity under different application monitoring scene, to be conducive to the realization of front end watch-dog product.
As shown in Figure 1, composition structural representation for video monitoring system, wherein video surveillance applications system roughly comprises video camera, front end video monitoring apparatus, transmission network and client device etc. end to end, to realize video acquisition, coding, Internet Transmission, decoding and to show the functions such as output.
As shown in Figure 2, be video coding process schematic diagram in video monitoring system, as shown in Figure 3, be Video coding framework schematic diagram detailed in video monitoring system.For the general coding flow process here, video monitoring, because application scenarios is comparatively complicated, can, for the feature of different application monitoring scenes, be optimized this coding scheme targetedly, to obtain the best video quality under application-specific scene, and reduce the complexity of encoding.
The monitoring scene that the embodiment of the present invention proposes here can be divided into static/moving scene, daytime/night scene, indoor/outdoor scene three classes, can adopt and carry out with the following method to distinguish respectively different monitoring scenes.
Embodiment mono-, the differentiation of motion/static scene:
As shown in Figure 4, the process chart that the differentiation video monitoring scene proposing for the embodiment of the present invention is motion/static scene, wherein can adopt motion detection to differentiate video monitoring scene is that moving scene is still static scene, the present frame of the video that is about to monitor and former frame image subtraction, obtain residual frame.As the not motion of front and back two two field picture correspondence positions, the correspondence position pixel brightness value (referred to as brightness value) of the residual frame obtaining is zero; As the not motion of front and back two two field picture correspondence positions, the brightness value that the correspondence position pixel brightness value of the residual frame obtaining is non-zero.The brightness value of each pixel of residual frame is added up, and as all pixel number object ratio values that comprise in the pixel number of non-zero brightness value and residual frame are greater than certain thresholding, differentiating current video monitoring scene is moving scene; Otherwise differentiating current video monitoring scene is static scene.Specific implementation flow process is as follows:
Accordingly, the embodiment of the present invention also proposes a kind of processing unit that video monitoring scene is motion/static scene of differentiating, the concrete composition structure of this device as shown in Figure 5, comprise that picture frame obtains unit 50, the last video frame image and the current video image frame that for obtaining video monitoring system front end video monitoring apparatus, monitor; Residual frame obtains unit 52, for picture frame being obtained to current video image frame that unit 50 obtains and last video frame image, subtracts each other and obtains Image Residual frame; Brightness value determining unit 54, for determining respectively the brightness value of each pixel of Image Residual frame that residual frame acquisition unit 52 obtains; Ratio value determining unit 56, for the brightness value of each pixel of determining respectively according to brightness value determining unit 54, determines the ratio value of the number of all pixels in the pixel number of non-zero brightness value and described residual frame; Scene judgement unit 58, for when the definite ratio value of ratio value determining unit 56 is greater than the proportion threshold value of setting, differentiating the current residing monitoring scene of front end video monitoring apparatus is moving scene, otherwise differentiate the current residing monitoring scene of front end video monitoring apparatus, is static scene.
Embodiment bis-, daytime/night scene differentiation:
As shown in Figure 6, the differentiation video monitoring scene proposing for the embodiment of the present invention is the process chart of scene at daytime/night, the embodiment of the present invention has represented image brightness based on brightness value in picture frame, by adding up the brightness value average of all pixels in a two field picture, the brightness value average obtaining as statistics is greater than certain thresholding, and can differentiate the residing monitoring scene of front end video monitoring apparatus is scene on daytime; The brightness value average obtaining as added up is lower than this thresholding, and can differentiate the residing monitoring scene of front end video monitoring apparatus is scene at night.Specific implementation flow process is as follows:
In addition, also can by the detection of time period is differentiated the scene of video monitoring be daytime scene or night scene, for example, within the time period of 6:00~19:00, can differentiate video monitoring scene is scene on daytime, is night monitoring scene in all the other time periods.
Accordingly, the embodiment of the present invention also proposes a kind of processing unit that video monitoring scene is scene at daytime/night of differentiating, the concrete composition structure of this device as shown in Figure 7, specifically comprise that picture frame obtains unit 70, the current video image frame monitoring for obtaining video monitoring system front end video monitoring apparatus; Brightness value average determining unit 72, for determining the average of the brightness value of each pixel that the current video image frame of picture frame acquisition unit 70 acquisitions comprises; Scene judgement unit 74, while being greater than the average threshold value of setting for the average at the definite brightness value of brightness value average determining unit, differentiating the current residing monitoring scene of front end video monitoring apparatus is scene on daytime; Otherwise differentiating the current residing monitoring scene of front end video monitoring apparatus is scene at night.
Embodiment tri-, the differentiation of indoor/outdoor scene:
As shown in Figure 8, the differentiation video monitoring scene proposing for the embodiment of the present invention is the process chart of indoor/outdoor scene, in embodiment of the present invention divided chamber, monitoring scene or outdoor monitoring scene are mainly paid close attention to whether uniform-illumination of video frame image, the general uneven illumination of video frame image wherein monitoring under Indoor Video scene is even, as impacts such as light.The piece that the two field picture monitoring is divided into 16x16 pixel (can certainly be divided into other big or small pieces, block of pixels such as the sizes such as 8 * 8,4 * 4), and calculate the brightness average Y of the block of pixels that each division obtains, obtain the high-high brightness average Y of block of pixels in a two field picture
maxwith minimum brightness average Y
min, as the Y obtaining
maxand Y
minmeet following formula, the scene of differentiating the current monitoring of front end video monitoring apparatus is indoor scene; Otherwise the scene of differentiating the current monitoring of front end video monitoring apparatus is outdoor scene.
Wherein TH can value be 2 conventionally.Its concrete handling process is as follows:
Accordingly, the embodiment of the present invention also proposes a kind of processing unit that video monitoring scene is indoor/outdoor scene of differentiating, the concrete composition structure of this device as shown in Figure 9, specifically comprises that picture frame obtains unit 90, the current video image frame monitoring for obtaining front end video monitoring apparatus; Piece division unit 92, is divided into the piece of M * N pixel size for picture frame being obtained to current video image frame that unit 90 obtains, wherein M, N are natural number; Brightness average determining unit 94, for the brightness average of determining that respectively piece division unit 92 is divided each piece obtaining; Minimax brightness value determining unit 96, for the brightness average of each piece of determining respectively in brightness average determining unit 94, determines maximum brightness average Y
maxwith minimum brightness average Y
min; Scene judgement unit 98, at the definite Y of minimax brightness value determining unit 96
maxand Y
minmeet
time, differentiating the current residing monitoring scene of front end video monitoring apparatus is indoor scene; Otherwise differentiating the current residing monitoring scene of front end video monitoring apparatus is outdoor scene, wherein TH is the quotient threshold value of setting.
The method of certain automatic discrimination monitoring scene except above introduction also can manually arrange monitoring scene pattern in front end video monitoring apparatus.
The embodiment of the present invention is differentiated scheme for the monitoring scene of above-mentioned proposition, proposes code optimization scheme targetedly respectively for every kind of different monitoring scene, specifically will describe in detail successively following.
Embodiment tetra-, the code optimization scheme under moving scene:
In video monitoring system, video acquisition situation when moving scene moves corresponding to video camera and individual soldier's equipment uses, static scene is the video acquisition situation fixedly time corresponding to video camera.
When moving scene, there is the motions such as taking turn in monitoring scene, estimation precision is lower due to video camera, and motion vector numerical value is very large, easily cause encoding code stream to change greatly, when fixed network bandwidth for transmission, easily occur the phenomenons such as packet loss greatly having reduced video encoding quality.For this defect, the present invention is that strength proposition is under moving scene at monitoring scene, increases overall motion estimation operation before the prediction processing in cataloged procedure, obtains global motion vector MV
g, can be used for representing the motion vector of video camera.Then carry out block-based estimation, obtain block motion vector MV
b, and calculate its difference MV
d:
MV
d=MV
b-MV
g (2)
Then with MV
dfor carrying out motion vector encoder subsequently in basis, to reduce the numerical values recited of motion vector, improve the quality of Video coding.
Therefore, as shown in figure 10, the first proposing for the embodiment of the present invention is for the process chart of the video monitoring image coding method of moving scene, and its specific implementation process is as follows:
Step 100, the current video image frame that the front end video monitoring apparatus in video monitoring system is monitored first carries out overall motion estimation before coded prediction, obtains global motion vector MV
g;
Step 101 is first carried out block-based estimation before coded prediction to current video image frame, obtain block motion vector MV
b;
Step 102, determines the block motion vector MV of above-mentioned acquisition
bwith global motion vector MV
gdifference MV
d;
Step 103, based on difference MV obtained above
dcurrent video image frame is carried out to motion vector encoder.
Accordingly, the embodiment of the present invention also proposes a kind of video monitoring image code processing apparatus that is moving scene for video monitoring scene, the concrete composition structure of this device as shown in figure 11, specifically comprise motion estimation unit 110, for the current video image frame that the front end video monitoring apparatus of video monitoring system is monitored, before coded prediction, carry out overall motion estimation, obtain global motion vector MV
g; And current video image frame is carried out before coded prediction to block-based estimation, obtain block motion vector MV
b; Vector difference determining unit 111, for the block motion vector MV that determines that motion estimation unit 110 obtains
bwith global motion vector MV
gdifference MV
d; Vector coding unit 112, for based on the definite difference MV of vector difference determining unit 111
dcurrent video image frame is carried out to motion vector encoder.
Further, while considering camera motion, may cause different images content frame to occur that the probability periodically repeating is less, therefore the reference frame quantity in the time of can considering to reduce Video coding, from multi-frame-reference, change single frames reference into, when not reducing precision of prediction, reduce the amount of calculation of estimation.Because video monitoring is higher to the quality requirement of single-frame images, when motion too violent and cause code check to increase when too fast, need to adjust Rate Control mode, can keep particularly QP substantially constant, by dynamic reducing frame per second, reach constant code rate, with the quality of bonding two field picture, specific embodiments as shown in figure 12.
Therefore, as shown in figure 13, the second proposing for the embodiment of the present invention is for the process chart of the video monitoring image coding method of moving scene, and its specific implementation process is as follows:
Step 131 is carried out block-based estimation before coded prediction to current video image frame, obtain block motion vector MV
b;
Step 135 is first carried out overall motion estimation before coded prediction to the current video image frame after above-mentioned frame per second adjustment, obtain global motion vector MV
g';
Step 136 is first carried out block-based estimation before coded prediction to the current video image frame after frame per second adjustment, obtain block motion vector MV
b';
Accordingly, the embodiment of the present invention also proposes the video monitoring image code processing apparatus that another kind is moving scene for video monitoring scene, the concrete composition structure of this device as shown in figure 14, specifically comprise motion estimation unit 141, for the current video image frame that the front end video monitoring apparatus of video monitoring system is monitored, before coded prediction, carry out overall motion estimation, obtain global motion vector MV
g; And current video image frame is carried out before coded prediction to block-based estimation, obtain block motion vector MV
b; Vector difference determining unit 143, for the block motion vector MV that determines that motion estimation unit 141 obtains
bwith global motion vector MV
gdifference MV
d; Motion vector encoder unit 145, for based on the definite difference MV of vector difference determining unit 143
dcurrent video image frame is carried out to motion vector encoder for the first time; Frame per second adjustment unit 147, for carrying out frame per second adjustment according to the coding result of motion vector encoder unit 145 to current video image frame; Subsequent motion estimation unit 141 also carries out overall motion estimation for the current video image frame to after the 145 frame per second adjustment of frame per second adjustment unit before coded prediction, obtains global motion vector MV
g'; And the current video image frame after frame per second adjustment is carried out to block-based estimation before coded prediction, obtain block motion vector MV
b'; Vector difference determining unit 143 is also for determining block motion vector MV
b' and global motion vector MV
g' difference MV
d'; Motion vector encoder unit 145 is also for based on difference MV
d' the current video image frame after frame per second adjustment is carried out to motion vector encoder for the second time.
Embodiment five, the code optimization scheme under static scene:
When monitoring scene is static scene, because the background parts of monitoring image remains unchanged, only when occurring, moving object occurs changing at fraction image-region, therefore can consider only the part of variation to be encoded, farthest to reduce encoder bit rate.As shown in figure 15, the process chart of the video monitoring image coding method for static scene proposing for the embodiment of the present invention, its specific implementation process is as follows:
Step 150, judges whether the current image frame that the front end video monitoring apparatus in video monitoring system monitors is the first frame video image; When judgment result is that while being, perform step 151; Otherwise perform step 152;
Based on above-mentioned proposition at monitoring scene, be the code optimization scheme under static scene, just can in the situation that keeping monitor video picture quality constant, greatly reduce the complexity that encoding and decoding are calculated.In addition because actual coding frame number reduces, and the coding region of coded frame reduces, and therefore in the situation that code check is constant, can improve the coding quality of each frame video image.
Accordingly, the embodiment of the present invention also proposes a kind of video monitoring image code processing apparatus that is static scene for video monitoring scene, the concrete composition structure of this device as shown in figure 16, specifically comprise residual frame determining unit 160, for current video image frame that the front end video monitoring apparatus of video monitoring system is monitored with respect to the region of variation as the first frame video image frame with reference to frame monitoring as residual frame; Video encoding unit 161, for the definite residual frame of described residual frame determining unit 160 is carried out to estimation with respect to the last video frame image monitoring, and carries out Video coding according to motion estimation result.
Embodiment six, the code optimization scheme under daytime/night scene:
Under daytime scene because illumination is strong, video acquisition image will be excessively bright, pixel brightness value is too concentrated, cause in cataloged procedure its ac coefficient AC after conversion process will be all close to zero, after quantizing, AC coefficient is directly zero, therefore can cause image detail to be lost, show as image a vast expanse of whiteness, indistinguishable details wherein.For scene at night, owing to gathering, image is excessively dark, also has the excessive a large amount of problems of losing of image detail that directly cause of identical quantization error.
For this problem, can carry out in coded quantization process at the current video image frame that front end video monitoring apparatus is monitored, the null value region obtaining after conversion is processed, reduce and quantize step value QStep
0, obtain new quantization step value QStep
0', the quantization step value QStep based on new then
0' current video image frame is encoded.Wherein:
QStep
0′=QStep
0-d (3)
Like this, at the video monitoring image under night on daytime scene, carry out in cataloged procedure, just can realize less conversion coefficient and be retained, and then retain more image detail, promoted the coding quality of image.In actual applications, quantize adjusting coefficient d, to can be, but not limited to value be QStep
0/ 2.
As shown in figure 17, the schematic diagram of for the embodiment of the present invention, null value area quantization being optimized and revised.
In addition, under scene at night, also can take identical null value area quantization method of adjustment to avoid the video image details monitoring seriously to lose.In addition due to the restriction of video camera technique, when gather video image night, can inevitably there are a large amount of noises, greatly affected the quality of the video image monitoring, also increased encoder bit rate, for addressing this problem, can before Image Coding, to current video image frame, carry out filtering processing to remove noise, promote the coding quality of video image.Under concrete night scene, code optimization schematic flow sheet as shown in figure 18.
Accordingly, it is the video monitoring image code processing apparatus of scene at night on daytime for video monitoring scene that the embodiment of the present invention also proposes a kind of, the concrete composition structure of this device as shown in figure 19, specifically comprise and reduce step value unit 190, for carrying out coded quantization process at the current video image frame that front end video monitoring apparatus is monitored, reduce and quantize step value; Video encoding unit 191, for based on reducing quantization step value after step value unit 190 reduces to the processing of encoding of current video image frame.Filter unit 192, for being that night is during scene at the current residing monitoring scene of front end video monitoring apparatus, quantization step value after reducing based on reduction step value unit is carried out filtering processing to current video image frame before current video image frame is encoded.
Embodiment seven, the code optimization scheme under indoor/outdoor scene:
When video monitoring scene is indoor scene, often, because the reasons such as light-illuminating produce the impact of brightness irregularities on video image, tend to exist the region of bright (around bulb) and excessively dark (corner that light is blocked).With aforementioned night on daytime scene the same, may lose due to the operation of uniform quantization a large amount of detail content of image.Therefore, for the video monitoring image frame under indoor scene, the every frame video image frame of current acquisition can be divided into the piece of M * N pixel size, wherein M, N are natural number, can be, but not limited to be divided into the piece of the pixel sizes such as 8 * 8,4 * 4,16 * 16,16 * 8 here.Each block of pixels obtaining based on division, can determine respectively the brightness average Y of each piece based on following formula:
Y wherein
ijfor the brightness value of each pixel of comprising in piece, i, j are positive integer.
According to the brightness average Y of each piece of determining, select brightness average to be less than the piece of the first setting threshold and the piece that brightness average is greater than the second setting threshold thereof, wherein the first setting threshold is less than the second setting threshold; The piece B of each M * N pixel size that soon this current video picture frame division will obtain is divided into two class B
uand B
n:
Wherein TH1 is the first setting threshold, and TH2 is the second setting threshold, and TH1 can get 64, TH2 and can get 191 conventionally.For the piece of above-mentioned selection (for B
upiece), can carry out the processing of null value area quantization, for the B choosing
upiece, in carrying out coded quantization process, by reducing its quantization step value QStep
0for QStep
0'; Then the quantization step value QStep based on after reducing
0' to the B selecting
upiece is encoded.And for unselected (for B
npiece), can be according to the processing of encoding of existing normal encoding mode.Such Optimized Coding Based mode can well promote bright area or cross the Image Coding quality of dark areas.
While being outdoor scene for video monitoring scene, also can be according to the processing of encoding of existing normal encoding mode.If under outdoor scene, be scene or night during scene on daytime simultaneously, can the Optimized Coding Based mode under scene carry out corresponding Optimized Coding Based processing in conjunction with daytime/night of above-mentioned by the agency of.
Accordingly, the embodiment of the present invention also proposes a kind of video monitoring image code processing apparatus that is indoor scene for video monitoring scene, the concrete composition structure of this device as shown in figure 20, specifically comprise that picture frame obtains unit 200, the current video image frame monitoring for obtaining the front end video monitoring apparatus of video monitoring system; Piece division unit 201, is divided into the piece of M * N pixel size for picture frame being obtained to current video image frame that unit 200 obtains, wherein M, N are natural number; Brightness average determining unit 202, for the brightness average of determining that respectively piece division unit 201 is divided each piece obtaining, wherein brightness average determining unit 202 can be according to formula
the brightness average of each piece that definite division obtains: Y wherein
ijfor the brightness value of each pixel of comprising in piece, i, j are positive integer; Piece selected cell 203, be used for according to the brightness average of each definite piece of brightness average determining unit 202, select brightness average to be less than the piece of the first setting threshold and the piece that brightness average is greater than the second setting threshold thereof, wherein the first setting threshold is less than the second setting threshold; Reduce step value unit 204, for the piece that piece selected cell 203 is selected, carry out coded quantization process, reduce and quantize step value; Video encoding unit 205, the piece of piece selected cell 203 being selected for the quantization step value based on reducing after step value unit 204 the reduces processing of encoding.
Arbitration schemes for aforementioned various identification monitoring scenes of having set forth respectively, and the Different Optimization encoding scheme proposing for the different monitoring scene recognizing, the embodiment of the present invention proposes a kind of new video monitoring coding method, specifically as shown in figure 21, the process chart of the video monitoring coding method proposing for the embodiment of the present invention, specific implementation process is as follows:
Step 210, differentiates the current residing monitoring scene of front end video monitoring apparatus in video monitoring system; Monitoring scene can be, but not limited to comprise motion/static scene, daytime/night scene, indoor/outdoor scene etc. particularly, wherein how to identify above-mentioned various monitoring scene in above-mentioned elaboration, no longer too much repeats here.
Step 211, according to the corresponding relation of predefined monitoring scene and Video coding mode, determines the Video coding mode that the above-mentioned monitoring scene determining is corresponding; Wherein under different monitoring scenes, can set in advance Optimized Coding Based mode that should monitoring scene, wherein for different monitoring scenes, adopt the coded system of which kind of optimization in above-mentioned elaboration, also no longer too much repeat here.
Step 212, every frame video image information front end video monitoring apparatus being monitored according to the above-mentioned Video coding mode of the determining processing of encoding.The specific coding process of concrete every kind of Optimized Coding Based mode also, in above-mentioned elaboration, also no longer too much repeats here.
Accordingly, the embodiment of the present invention also proposes a kind of video monitoring code processing apparatus, the concrete composition structure of this device as shown in figure 22, specifically comprises scene judgement unit 221, for differentiating the current residing monitoring scene of front end video monitoring apparatus of video monitoring system; Coded system determining unit 222, for according to the corresponding relation of predefined monitoring scene and Video coding mode, determines Video coding mode corresponding to monitoring scene that scene judgement unit 221 determines; Coding unit 223, every frame video image information front end video monitoring apparatus being monitored for the Video coding mode of determining according to coded system determining unit 222 processing of encoding.How how scene judgement unit 221 identify above-mentioned various monitoring scene, coded system determining unit 222 for different monitoring scenes particularly, adopt the coded system of which kind of optimization and specific coding process that coding unit 223 adopts every kind of Optimized Coding Based mode particularly all to distinguish in above-mentioned elaboration, no longer too much repeat here.
To sum up discuss visible, the embodiment of the present invention is by proposing the identifying schemes of different monitoring scene, and for the different monitoring scene recognizing, the adaptive encoding scheme of monitoring image is targetedly proposed, thereby realized for different monitoring scenes and carried out respectively image optimization coding, promote the video monitoring image quality under various monitoring scenes, reduced the complexity of encoding video pictures, and then effectively promoted the implementation result of Video Supervision Technique.
Obviously, those skilled in the art can carry out various changes and modification and not depart from the spirit and scope of the present invention the present invention.Like this, if within of the present invention these are revised and modification belongs to the scope of the claims in the present invention and equivalent technologies thereof, the present invention is also intended to comprise these changes and modification interior.
Claims (2)
1. a video monitoring scene method of discrimination, is characterized in that, comprising:
Obtain last video frame image and current video image frame that front end video monitoring apparatus monitors;
The current video image frame of acquisition and last video frame image are subtracted each other and obtain Image Residual frame;
Determine the brightness value of each pixel in the Image Residual frame obtaining;
According to the brightness value of each pixel of determining, determine the ratio value of the number of all pixels in the pixel number of non-zero brightness value and described residual frame;
If definite ratio value is greater than the proportion threshold value of setting, differentiating the current residing monitoring scene of front end video monitoring apparatus is moving scene; Otherwise
Differentiating the current residing monitoring scene of front end video monitoring apparatus is static scene.
2. a video monitoring scene discriminating gear, is characterized in that, comprising:
Picture frame obtains unit, the last video frame image and the current video image frame that for obtaining front end video monitoring apparatus, monitor;
Residual frame obtains unit, for picture frame being obtained to current video image frame that unit obtains and last video frame image, subtracts each other and obtains Image Residual frame;
Brightness value determining unit, for determining the brightness value of each pixel of Image Residual frame that residual frame acquisition unit obtains;
Ratio value determining unit, for the brightness value of each pixel of determining according to brightness value determining unit, determines the ratio value of the number of all pixels in the pixel number of non-zero brightness value and described residual frame;
Scene judgement unit, for when the definite ratio value of ratio value determining unit is greater than the proportion threshold value of setting, differentiating the current residing monitoring scene of front end video monitoring apparatus is moving scene; Otherwise differentiating the current residing monitoring scene of front end video monitoring apparatus is static scene.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310491808.XA CN103561232A (en) | 2010-12-29 | 2010-12-29 | Method and device for identifying video monitoring scenes |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310491808.XA CN103561232A (en) | 2010-12-29 | 2010-12-29 | Method and device for identifying video monitoring scenes |
Related Parent Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN2010106123774A Division CN102572381A (en) | 2010-12-29 | 2010-12-29 | Video monitoring scene judging method and monitoring image coding method and device thereof |
Publications (1)
Publication Number | Publication Date |
---|---|
CN103561232A true CN103561232A (en) | 2014-02-05 |
Family
ID=50015371
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310491808.XA Pending CN103561232A (en) | 2010-12-29 | 2010-12-29 | Method and device for identifying video monitoring scenes |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103561232A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105744270A (en) * | 2014-12-10 | 2016-07-06 | 中国移动通信集团公司 | Cloud-platform-based video coding method and apparatus |
CN106604029A (en) * | 2017-01-03 | 2017-04-26 | 华南理工大学 | HEVC-based bit rate control method for motion region detection |
CN109361923A (en) * | 2018-12-04 | 2019-02-19 | 深圳市梦网百科信息技术有限公司 | A kind of time slip-window scene change detection method and system based on motion analysis |
CN113163194A (en) * | 2021-02-05 | 2021-07-23 | 珠海视熙科技有限公司 | Detection method and device for flicker of shooting scene, shooting equipment and medium |
CN114339431A (en) * | 2021-12-16 | 2022-04-12 | 杭州当虹科技股份有限公司 | Time-lapse coding compression method |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH09233422A (en) * | 1996-02-26 | 1997-09-05 | Oki Electric Ind Co Ltd | Moving image information detector in moving image processing system |
US20070098084A1 (en) * | 2005-10-31 | 2007-05-03 | Fujitsu Limited | Moving picture encoder |
US20070098074A1 (en) * | 2005-10-31 | 2007-05-03 | Fujitsu Limited | Moving picture encoding device, fade scene detection device and storage medium |
CN101017573A (en) * | 2007-02-09 | 2007-08-15 | 南京大学 | Method for detecting and identifying moving target based on video monitoring |
CN101052131A (en) * | 2006-04-06 | 2007-10-10 | 株式会社尼康 | Imaging apparatus |
CN101764922A (en) * | 2009-08-03 | 2010-06-30 | 北京智安邦科技有限公司 | Method and device for adaptive generation of luminance threshold |
CN101872526A (en) * | 2010-06-01 | 2010-10-27 | 重庆市海普软件产业有限公司 | Smoke and fire intelligent identification method based on programmable photographing technology |
-
2010
- 2010-12-29 CN CN201310491808.XA patent/CN103561232A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH09233422A (en) * | 1996-02-26 | 1997-09-05 | Oki Electric Ind Co Ltd | Moving image information detector in moving image processing system |
US20070098084A1 (en) * | 2005-10-31 | 2007-05-03 | Fujitsu Limited | Moving picture encoder |
US20070098074A1 (en) * | 2005-10-31 | 2007-05-03 | Fujitsu Limited | Moving picture encoding device, fade scene detection device and storage medium |
CN101052131A (en) * | 2006-04-06 | 2007-10-10 | 株式会社尼康 | Imaging apparatus |
CN101017573A (en) * | 2007-02-09 | 2007-08-15 | 南京大学 | Method for detecting and identifying moving target based on video monitoring |
CN101764922A (en) * | 2009-08-03 | 2010-06-30 | 北京智安邦科技有限公司 | Method and device for adaptive generation of luminance threshold |
CN101872526A (en) * | 2010-06-01 | 2010-10-27 | 重庆市海普软件产业有限公司 | Smoke and fire intelligent identification method based on programmable photographing technology |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105744270A (en) * | 2014-12-10 | 2016-07-06 | 中国移动通信集团公司 | Cloud-platform-based video coding method and apparatus |
CN105744270B (en) * | 2014-12-10 | 2019-01-01 | 中国移动通信集团公司 | A kind of method for video coding and device with holder |
CN106604029A (en) * | 2017-01-03 | 2017-04-26 | 华南理工大学 | HEVC-based bit rate control method for motion region detection |
CN106604029B (en) * | 2017-01-03 | 2019-07-12 | 华南理工大学 | A kind of bit rate control method of the moving region detection based on HEVC |
CN109361923A (en) * | 2018-12-04 | 2019-02-19 | 深圳市梦网百科信息技术有限公司 | A kind of time slip-window scene change detection method and system based on motion analysis |
CN109361923B (en) * | 2018-12-04 | 2022-05-31 | 深圳市梦网视讯有限公司 | Sliding time window scene switching detection method and system based on motion analysis |
CN113163194A (en) * | 2021-02-05 | 2021-07-23 | 珠海视熙科技有限公司 | Detection method and device for flicker of shooting scene, shooting equipment and medium |
CN114339431A (en) * | 2021-12-16 | 2022-04-12 | 杭州当虹科技股份有限公司 | Time-lapse coding compression method |
CN114339431B (en) * | 2021-12-16 | 2023-09-01 | 杭州当虹科技股份有限公司 | Time-lapse coding compression method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102547225B (en) | Video monitoring scene judgment method and device, and monitored image encoding method and device | |
CN102572380B (en) | Video monitoring coding method and device | |
CN102572381A (en) | Video monitoring scene judging method and monitoring image coding method and device thereof | |
CN102395030B (en) | Motion analysis method based on video compression code stream, code stream conversion method and apparatus thereof | |
CN102014279B (en) | Method and device for intensifying video image contrast | |
CN103561232A (en) | Method and device for identifying video monitoring scenes | |
CN102420988B (en) | Multi-view video coding system utilizing visual characteristics | |
CN106296612A (en) | The stagewise monitor video sharpening system and method that a kind of image quality evaluation and weather conditions guide | |
CN103179325A (en) | Self-adaptive 3D (Three-Dimensional) noise reduction method for low signal-to-noise ratio video under fixed scene | |
CN104023166B (en) | A kind of environment self-adaption video image noise reducing method and device | |
CN104954791A (en) | Method for selecting key frame from wireless distributed video coding for mine in real time | |
CN105072345A (en) | Video encoding method and device | |
CN103079062A (en) | Intelligent video surveillance system | |
CN110662044A (en) | Video coding method, video coding device and computer storage medium | |
CN103813176A (en) | Deblocking filter method and adaptive loop filter method in video encoding and decoding | |
CN102014243A (en) | Method and device for enhancing images | |
CN105451016A (en) | No-reference video quality evaluation method suitable for video monitoring system | |
CN104202599B (en) | A kind of all-zero blocks method applied to HEVC interframe encodes | |
CN101945281B (en) | Method and device for filtering video codes | |
CN107147912B (en) | Video coding method and device | |
CN107257464A (en) | HD video encoder bit rate control algolithm based on Sobel operators and linear regression | |
CN103561268A (en) | Method and device for encoding video monitoring image | |
CN103747257B (en) | A kind of method of video data high efficient coding | |
CN102663682A (en) | Adaptive image enhancement method based on interesting area | |
CN101540834A (en) | Method for removing noise of video image and video coding device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20140205 |