The content of the invention
In view of this, the embodiment of the present invention provides a kind of new distributed video analysis method and system, solves existing
Headend equipment or the problem of the video analysis limited ability of rear end equipment in technology, can be realized under limited device resource configuration
The video analysis of more big data quantity.
A kind of distributed video analysis method provided in an embodiment of the present invention, includes at headend equipment:
Obtain the fundamental analysis result of video to be analyzed;
The video image of the fundamental analysis result and the video to be analyzed is set up into corresponding relation;
The fundamental analysis result, the corresponding relation and the video formation mixed data flow to be analyzed are sent to
Rear end equipment.
A kind of distributed video analysis system provided in an embodiment of the present invention includes:At least one headend equipment;Wherein, institute
Stating at least one headend equipment includes:
Fundamental analysis module, is configured to obtain the fundamental analysis result of video to be analyzed;
Respective modules, are configured to fundamental analysis result pass corresponding with the video image foundation of the video to be analyzed
System;
Mixing module, is configured to fundamental analysis result, the corresponding relation and the video formation mixing to be analyzed
Data flow.
The distributed video analysis method provided using the embodiment of the present invention, headend equipment utilizes simple computing function
Fundamental analysis result is obtained, the corresponding relation of fundamental analysis result and video to be analyzed is set up, and by fundamental analysis result and right
It should be related to and be sent to rear end equipment.Rear end equipment directly can further complete more multiple using the fundamental analysis result and corresponding relation
Miscellaneous video analysis.So headend equipment and rear end equipment all need not very big amount of calculation, realize headend equipment and rear end
The balance of equipment computing capability, so as to improve the efficiency of video analysis so that limited device resource configuration can be realized bigger
The video analysis of data volume.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation is described, it is clear that described embodiment is only a part of embodiment of the invention, rather than whole embodiments.It is based on
Embodiment in the present invention, it is every other that those of ordinary skill in the art are obtained under the premise of creative work is not made
Embodiment, belongs to the scope of protection of the invention.
A kind of distributed video analysis method provided in an embodiment of the present invention, sets applied to one kind including at least one front end
The distributed video analysis system of standby and at least one rear end equipment;Wherein, fundamental analysis is carried out by least one headend equipment,
Complicated analysis is carried out by least one rear end equipment.The fundamental analysis of headend equipment is illustrated by embodiment separately below
The complicated analysis process of process and rear end equipment.
Fig. 1 show the fundamental analysis of headend equipment in the distributed video analysis method that one embodiment of the invention is provided
Schematic flow sheet.As shown in figure 1, the fundamental analysis includes:
Step 101:Obtain the fundamental analysis result of video to be analyzed.
In an embodiment of the present invention, can carry out base for some in video image or multiple objects to be analyzed
Plinth is analyzed, and object to be analyzed can be prospect, moving target or other specific objectives (face, car plate) in video image etc..
The purpose of fundamental analysis is the fundamental analysis information for obtaining object to be analyzed.The fundamental analysis information generally contains
The important core information of object to be analyzed, has abandoned the detailed description information of object to be analyzed.For example, when object to be analyzed
During for " car plate ", fundamental analysis information can only include digital profile information in car plate, without the color comprising car plate and
Positional information.
The fundamental analysis information of object to be analyzed can be:Position, size, shape, profile and the color of object to be analyzed
Information etc..It will be understood by those skilled in the art that user can determine fundamental analysis information according to the attribute of object to be analyzed
Content.The particular content of fundamental analysis information can not be used to limit the scope of the invention.
Step 102:The video image of fundamental analysis result and video to be analyzed is set up into corresponding relation.
In actual applications, because object to be analyzed not necessarily appears in all video figures of video to be analyzed all the time
As in, some video images do not include object to be analyzed, thus will not also produce fundamental analysis result.Therefore, one it is to be analyzed
The fundamental analysis result of video may also come from multiple discontinuous video images of the video to be analyzed, or be treated from one
Certain width video image of analysis video can obtain multiple fundamental analysis results;By by fundamental analysis result and video to be analyzed
Video image correspondence, can be easy to quickly position the video image for including object to be analyzed during complicated analysis below,
So as to improve the efficiency of follow-up complicated analysis process.
It will be understood by those skilled in the art that following coded system can be used by setting up the mode of corresponding relation:
Each fundamental analysis result, each video image and each object to be analyzed are numbered respectively;
It will be analyzed based on the numbering and object number to be analyzed of the corresponding video image of each fundamental analysis result
As a result property value is recorded;Or
It regard the numbering and object number to be analyzed of the corresponding fundamental analysis result of each video image as video image
Property value recorded.
Step 103:Fundamental analysis result, corresponding relation and video to be analyzed formation mixed data flow are sent at least
One rear end equipment.Fundamental analysis result is formed into mixed data flow in the form of standardizing with video to be analyzed.
In an embodiment of the present invention, it then follows《Safety precaution monitoring digital video-audio encoding and decoding technique requires GB/T
25724-2010》Monitoring in (hereinafter referred to as SVAC standards) extends the definition of information to represent fundamental analysis result.Wherein
The grammer content of monitoring extension information in SVAC standards may include one or more of following items:Extension information type,
Analysis result subtype, corresponding video frame number, corresponding frame of video absolute time information, extension information data length, pressure
Reduce the staff coded data of yard algorithm types and analysis result etc..Surveillance center's termination receives the mixed number for including fundamental analysis result
After stream, the growth data for including fundamental analysis result is directly extracted, some of video can be obtained without decoding video images
Fundamental analysis result, and can be realized according to the information in fundamental analysis result, such as target location, size, type, color
The quick-searching of video image is played.
So rear end equipment can directly extract existing fundamental analysis result and be answered when receiving the mixed data flow
Miscellaneous analyzing and processing, without analyzing again the video image of video to be analyzed, reduces the amount of calculation of rear end equipment, shows
Write the processing pressure for reducing rear end equipment so that the server equally configured can complete the analyzing and processing of more multi-channel video.
When the data volume needed for the content of fundamental analysis result is unable to reach complicated analysis, rear end equipment can also be utilized
Fundamental analysis result and the corresponding relation quick-searching of the video image of video to be analyzed go out the video figure comprising object to be analyzed
Picture, and complicated analysis is only carried out to the object to be analyzed in the video image that retrieves, so as to improve the efficiency of complicated analysis.
In an embodiment of the present invention, can also be to basis point after the fundamental analysis result of video to be analyzed is obtained
Analysis result is compressed, and same fundamental analysis information is represented with less data volume.
It will be understood by those skilled in the art that user can be according to the data characteristicses of fundamental analysis result, selection is suitable
Data compression scheme, but the selection of compress mode can not be used to limit the scope of the invention.
Such as, when to obtain the profile information of object to be analyzed, it is to be analyzed right can be included in fundamental analysis result
The video image of the video to be analyzed of elephant is converted into binary image, prospect (object to be analyzed) is such as represented with " 1 ", with " 0 " generation
Table background;Data compression is carried out to binary image using Run- Length Coding (RLC, Run Length Coding) again.So not only
The size of data of fundamental analysis result can be substantially reduced, the profile information of object to be analyzed can also be highlighted.
In an embodiment of the present invention, mixed data flow can be stored in headend equipment and/or rear end equipment.So user can
The mixed data flow is transferred from headend equipment and/or rear end equipment at any time, video analysis data is reviewed or by other to facilitate
Analytical equipment carries out further video analysis.Certainly, it is contemplated that the memory space requirement of mixed data flow is big, generally should
It is stored in rear end equipment.
Fig. 2 show the complicated analysis of rear end equipment in the distributed video analysis method that one embodiment of the invention is provided
Process flow schematic diagram.As shown in Fig. 2 the complicated analysis process includes:
Step 201:Extract the fundamental analysis result in mixed data flow.
Step 202:Judge whether the fundamental analysis result extracted can meet the video analysis demand of user, if basic
Analysis result disclosure satisfy that the video analysis demand of user, then directly tied fundamental analysis result as final complicated analysis
Really, so far flow terminates;Otherwise step 203 is performed.
Step 203:Complicated analysis is carried out to fundamental analysis result.Specifically, complicated point is carried out to fundamental analysis result
The process of analysis can be:
Filtered out using the corresponding relation of fundamental analysis result and the video image of video to be analyzed comprising object to be analyzed
Video image;
Complicated analysis is carried out to object to be analyzed for the video image screened.
So rear end equipment is just without analyzing the video image for not including object to be analyzed, so as to substantially increase
The complicated analysis efficiency of rear end equipment.
Wherein, according to the difference of object to be analyzed, complexity analysis may include:Behavioural analysis to object to be analyzed, and/or
Gripper path analysis, and/or Object identifying.
Fig. 3 show the structural representation for the distributed video analysis system that one embodiment of the invention is provided.Such as Fig. 3 institutes
Show, the system includes:At least one headend equipment and at least one rear end equipment;Wherein, headend equipment includes:
Fundamental analysis module, is configured to obtain the fundamental analysis result of video to be analyzed;
Respective modules, are configured to the video image of fundamental analysis result and video to be analyzed setting up corresponding relation;
Mixing module, is configured to fundamental analysis result, corresponding relation and video to be analyzed formation mixed data flow.
Compression module, is configured to compress fundamental analysis result.
Wherein, rear end equipment includes:
Extraction module, is configured to extract the fundamental analysis result in mixed data flow;
Screening module, the corresponding relation for being configured to the video image using fundamental analysis result and video to be analyzed is filtered out
Include the video image of object to be analyzed;
Complicated analysis module, is configured to carry out object to be analyzed complicated analysis for the video image filtered out.
In an embodiment of the present invention, it can further include in headend equipment and/or rear end equipment:
Memory module, is configured to storage fundamental analysis result and/or mixed data flow.
In an embodiment of the present invention, rear end equipment can also be configured to original fundamental analysis result and carry out complexity
Analysis.Now because fundamental analysis result is without being compressed, headend equipment can not include compression module yet.
It will be understood by those skilled in the art that according to the complexity and headend equipment of the video analysis to be realized with after
The computing capability of end equipment, user can adjust headend equipment and the respective number of rear end equipment in distributed video analysis system
Amount.The present invention is to the quantity of headend equipment and rear end equipment and is not construed as limiting.
These are only presently preferred embodiments of the present invention, be not intended to limit the invention, it is all the present invention spirit and
Within principle, any modification, equivalent substitution for being made etc. should be included in the scope of the protection.