CN110213651A - A kind of intelligent merit Computer Aided Analysis System and method based on security protection video - Google Patents
A kind of intelligent merit Computer Aided Analysis System and method based on security protection video Download PDFInfo
- Publication number
- CN110213651A CN110213651A CN201910449893.0A CN201910449893A CN110213651A CN 110213651 A CN110213651 A CN 110213651A CN 201910449893 A CN201910449893 A CN 201910449893A CN 110213651 A CN110213651 A CN 110213651A
- Authority
- CN
- China
- Prior art keywords
- video
- merit
- module
- analysis
- auxiliary
- 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
- 238000004458 analytical method Methods 0.000 title claims abstract description 111
- 238000000034 method Methods 0.000 title claims description 23
- 238000001514 detection method Methods 0.000 claims abstract description 65
- 230000003993 interaction Effects 0.000 claims abstract description 24
- 230000006870 function Effects 0.000 claims description 14
- 230000008569 process Effects 0.000 claims description 13
- 238000003058 natural language processing Methods 0.000 claims description 12
- 238000000513 principal component analysis Methods 0.000 claims description 9
- 238000001228 spectrum Methods 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 8
- 238000000605 extraction Methods 0.000 claims description 6
- 238000004519 manufacturing process Methods 0.000 claims description 6
- 238000012986 modification Methods 0.000 claims description 4
- 230000004048 modification Effects 0.000 claims description 4
- 230000004044 response Effects 0.000 claims description 4
- 238000000926 separation method Methods 0.000 claims description 4
- 230000009471 action Effects 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 abstract description 5
- 238000012544 monitoring process Methods 0.000 description 7
- 230000002452 interceptive effect Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 239000000284 extract Substances 0.000 description 3
- 238000010276 construction Methods 0.000 description 2
- 238000011835 investigation Methods 0.000 description 2
- 238000003860 storage Methods 0.000 description 2
- 230000001960 triggered effect Effects 0.000 description 2
- 241000406668 Loxodonta cyclotis Species 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005728 strengthening Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/43—Processing of content or additional data, e.g. demultiplexing additional data from a digital video stream; Elementary client operations, e.g. monitoring of home network or synchronising decoder's clock; Client middleware
- H04N21/44—Processing of video elementary streams, e.g. splicing a video clip retrieved from local storage with an incoming video stream or rendering scenes according to encoded video stream scene graphs
- H04N21/44008—Processing of video elementary streams, e.g. splicing a video clip retrieved from local storage with an incoming video stream or rendering scenes according to encoded video stream scene graphs involving operations for analysing video streams, e.g. detecting features or characteristics in the video stream
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/80—Generation or processing of content or additional data by content creator independently of the distribution process; Content per se
- H04N21/85—Assembly of content; Generation of multimedia applications
- H04N21/854—Content authoring
- H04N21/8549—Creating video summaries, e.g. movie trailer
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30241—Trajectory
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Signal Processing (AREA)
- Tourism & Hospitality (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Databases & Information Systems (AREA)
- Computer Security & Cryptography (AREA)
- Educational Administration (AREA)
- Development Economics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
A kind of intelligent merit Computer Aided Analysis System based on security protection video, including following module: the video file connecting with video source uploads and real-time streams AM access module, and the video source provides the video resource for assisting merit to analyze；Target detection and object tracking module；Video analysis algoritic module；Video summary generation module；Merit management and auxiliary video figure detect module；Human-computer interaction module；Merit manages database；Picture snapshot and video frequency abstract index database；Original video data library.A kind of intelligent merit aided analysis method based on security protection video, using a kind of above-mentioned intelligent merit Computer Aided Analysis System based on security protection video.The present invention have many advantages, such as auxiliary accelerate merit analysis, install it is simple, easy to use, economical and practical.The invention belongs to information technology fields.
Description
Technical field
The invention belongs to information technology field more particularly to a kind of intelligent merit Computer Aided Analysis Systems based on security protection video
And method.
Background technique
Since the Ministry of Public Security disposes " 3111 engineering project " (i.e. " safe city ") Pilot project construction with " science and technology strengthening police " project,
Public security video monitoring system has substantially been built in national prefectural level city.The construction of public security video monitoring system is public security machine
It closes and realizes that " fast reaction, precision strike, comprehensively control " target to social security state provides strong science and technology support, it is special
It is not to provide video information (real time information and subsequent video recording) by video monitoring system to have cracked a large amount of criminal case.At present
Public security video monitoring system has become police criminal detection and suspect inquires an indispensable technological means in work.Currently,
By there are mainly two types of the methods of security protection video resource auxiliary merit analysis: the first is auxiliary in such a way that manual video plays back
Case investigation and information is helped to study and judge, the mode that this mode relies primarily on artificial browsing video is studied and judged, there are low efficiency, easily
The disadvantages of omitting, be error-prone, but it is still current city, case investigation at county level and the main means studied and judged；Second is to pass through structure
Video structural big data platform is built, it is each to integrate security monitoring, public security bayonet, traffic police's monitoring, electronic police and intelligent bayonet etc.
Class video monitoring resource extracts the target object and its motion profile in video, and these targets by video analysis means
It is different classes of to support video frequency searching function to be categorized further, as people, vehicle, building, bicycle etc..This kind of system platform is usual
It is deployed in city, public security system Intranet above the provincial level, deployment is complicated, cost is very high, is unfavorable for following rank public security bureau, county, sends
The development at the basic level applications such as institute, cause that operation threshold is high, popularity rate is very low, effect is general.
Summary of the invention
In view of the above-mentioned problems, the present invention provides a kind of intelligent merit Computer Aided Analysis System based on security protection video, have
The advantages that merit assistant analysis is simpler, more convenient, more effective.
The second object of the present invention is to provide a kind of intelligent merit aided analysis method based on security protection video.
A kind of intelligent merit Computer Aided Analysis System based on security protection video, the video file including connecting with video source upload
With real-time streams AM access module, the video source provides the video resource for assisting merit to analyze,
For the target and the target detection that effectively extracts of motion profile and object tracking module in video resource,
The video analysis algoritic modules of all kinds of video analysis algorithms configured are executed for various,
For synthetic video as needed abstract video summary generation module,
The merit management of integrated management is carried out to the case of institute's assistant analysis and auxiliary video figure detects module,
For realizing the human-computer interaction module of the interactive operation of merit analyst and this system,
Merit for storing information data related with merit manages database,
For efficient storage target object, characteristic information, summarized radio, the picture snapshot of motion profile and video frequency abstract rope
Draw library,
For saving the original video data library of original video files related with case or live video stream；
Human-computer interaction module calls merit management and auxiliary video figure to detect module after the order for receiving merit analyst；
Merit management database is associated with picture snapshot and video frequency abstract index database, picture snapshot and video frequency abstract index database and original view
Frequency database association；
Merit management and auxiliary video figure detect module and call video analysis algoritic module, video summary generation module, video
Parser module, the detection of video summary generation module invocation target and object tracking module, target detection and to image tracing mould
Block analyzes original video library, and video summary generation module, target detection and object tracking module will generate result and protect
It deposits to picture snapshot and video frequency abstract index database, early warning is back to merit management and auxiliary video figure by video analysis algoritic module
Detect module；
Merit management and auxiliary video figure detect module polls merit management database, and merit manages database for query result
It is back to merit management and auxiliary video figure detects module, merit management database simultaneously saves inquiry record；
Merit management and auxiliary video figure, which detect module, will analyze result return human-computer interaction module.
As a preference, video source includes video file and video flowing, video file includes AVI, MPEG4, WMV lattice
Formula, video flowing include the live video stream that IP camera/DVR/NVR is provided.
As a preference, video file uploads and real-time streams AM access module stores the video file of upload to original view
Frequency database, and live video stream is saved in original by the MPEG4 video format file that video format coding changes into standard again
Beginning video database.
As a preference, target detection and object tracking module are by multi-target detection and track algorithm to video file
Or live video stream is analyzed, is handled, so that it is fast to export the size of target object, shape, color related information, picture
According to, motion profile, and it is saved into picture snapshot and video frequency abstract index database.
As a preference, multi-target detection and track algorithm include being based on multifractal spectra (Multi-Fractal
Spectrum, MFS) and Robust Principal Component Analysis (Robust Principal Component Analysis, RPCA), it realizes
Processing is analyzed simultaneously in the analytic process to video source, while multiple target objects, the respective motion profile of target object.
As a preference, being flexibly implanted into all kinds of intelligence by configuring on the basis of the multi-target detection and track algorithm
Energy video analysis algorithm, such as: described to be configured in intrusion detection, classification and Detection, color detection and crowd massing detection algorithm
Any one or a few, is arranged the trigger condition of video analysis algorithm, when the triggering of the trigger condition of video analysis algorithm, driving
The process action of merit assistant analysis.
As a preference, video analysis algoritic module and video summary generation module comprise the steps of:
(1) it receives instruction: receiving merit management and auxiliary video figure frame module is assigned and plucked for video source analysis and video
The order to be made；
(2) video file or live video stream are linked into target detection and object tracking module, and the module is first original
Beginning video file is stored in original video data library, and live video stream then encodes it, synchronizes and is converted to certain view
Frequency file format is stored in original video data library.
(3) file uploaded to video source carries out key-frame extraction by way of uniformly extracting；
(4) prospect is carried out to key frame, background separation operates；
(5) motion profile, the target pair of target object, target object are extracted in invocation target detection with object tracking module
Relevant information, the picture snapshot of target object of elephant；
(6) the video frequency abstract manufacturing conditions configured according to merit management and auxiliary video figure frame module, to specific mesh
Object is marked, the time response of the motion profile of pipe is fulfilled according to target, required video frequency abstract file is generated, these videos is plucked
It wants file to be written in picture snapshot and video frequency abstract index database, it is fast further to establish video source, Moving Objects, object picture
According to, the index between motion profile, video frequency abstract file, for subsequent retrieval operations use；
(7) it is finished if this video source has been analyzed, terminates this operating process；Otherwise, continue to repeat step (3) to step
Suddenly (6), until video source analysis finishes.
As a preference, human-computer interaction module includes: that (1) common many condition selection input or conditional attribute are defeated
Enter；(2) natural language text inputs.
As a preference, merit management and auxiliary video figure frame module include the common management function of merit and are based on regarding
The auxiliary figure of frequency analysis detects function；The common management function include establish case, modification case, delete case, inquiry case,
Upload case attachment；It is detectd in function in the auxiliary figure, by carrying out video analysis to video source related with merit, video is plucked
It generates, the video content and structural information of video is obtained, then by establishing between merit and video content, structural information
Association, for assisting merit to analyze.
A kind of intelligent merit aided analysis method based on security protection video, using a kind of above-mentioned intelligence based on security protection video
Merit Computer Aided Analysis System, includes the following steps:
(1) merit analyst inputted on human-computer interaction interface the alternative condition of general query text or multiple fixations into
Row inquiry, if it is general query text, needs to carry out semantic understanding, then natural language processing (NLP) algorithm is called to look into this
It askes text and realizes semantic understanding, and inquiry request is mapped as suitable relational database SQL statement sequence；Otherwise, direct root
It is investigated that asking the association that alternative condition establishes itself and SQL statement sequence；
(3) based on the SQL statement sequence in step (2), inquiry merit manages database, obtains video relevant with merit
Source mark；
(4) according to video source mark, in conjunction with the querying condition of the SQL statement in step (2), retrieving image snapshot and video
Abstract index database, to obtain picture snapshot, the video frequency abstract segment of qualified suspected target；These contents can be to scheme simultaneously
The mode of shape is shown to human-computer interaction interface, and the auxiliary figure that altar table mutual affection analysis personnel do next step detects decision；
(5) merit analyst is according to the picture snapshot of the suspected target of step (4), characteristic information, video frequency abstract segment, and
In conjunction with original video, auxiliary merit analysis is carried out；
(6) merit analyst combines merit situation, to suspicion object, calibration, filing-up work after carrying out analysis decision, and
Corresponding result is saved in merit management database and corresponds to case file；
(7) merit analysis personnel continue auxiliary figure if necessary and detect operation, are further continued for repeating the above steps (2) to step
Suddenly (6), if it is determined that terminating, then process terminates.
Beneficial effects of the present invention:
1, the present invention is centered on " merit ", with video file and IP camera etc. for main video source, target object
Detection combines to new and high technologies such as image tracing, video analysis algorithm, video frequency abstract generations together, passes through foundation " merit, suspicion
Incidence relation between Target Photo snapshot, suspected target summarized radio, original video ", the merit for overcoming manual video to play back
The corresponding drawback of assistant analysis means accelerates merit analysis with auxiliary；Also, the system is one-of-a-kind system, does not need more meters
It calculates machine equipment to support, therefore installation is simple, easy to use, economical and practical, is more suitable for public security bureau, the local police station of more base.
2, target detection and object tracking algorithm use based on multifractal spectra (Multi-Fractal Spectrum,
MFS) and the multi-target detection of Robust Principal Component Analysis (Robust Principal Component Analysis, RPCA) and
Multiple target objects, their respective motion profiles can be analyzed processing simultaneously by track algorithm, to promote Objective extraction and right
The efficiency of image tracing.
3, human-computer interaction module of the invention is changed existing by supporting natural language processing (NLP) and many condition to input
The interactive mode for having scheme that text many condition is only supported to input improves the user experience and merit of merit analyst
Integrated management ability and analysis efficiency.
4, the present invention analyses in depth " merit, suspected target picture snapshot, suspected target abstract view centered on " merit "
Frequently, the direct incidence relation of original video, intelligent analysis process, merit analysis workflow ", establishes the intelligence based on security protection video
It can merit assistant analysis working model, further promotion merit analysis efficiency and accuracy.
Detailed description of the invention
Fig. 1 is a kind of schematic diagram of intelligent merit Computer Aided Analysis System based on security protection video.
Fig. 2 is a kind of flow diagram of intelligent merit aided analysis method based on security protection video.
Fig. 3 is the stream of the video analysis and video frequency abstract in a kind of intelligent merit Computer Aided Analysis System based on security protection video
Journey schematic diagram.
Specific embodiment
The present invention is further illustrated with reference to the accompanying drawing.
As shown in Figure 1, a kind of intelligent merit Computer Aided Analysis System based on security protection video, it includes that file uploads and in real time
Flow AM access module, target detection and object tracking module, video analysis algoritic module, video summary generation module, merit management
Module, human-computer interaction module, merit management database, picture snapshot and video frequency abstract index database and original are detectd with auxiliary video figure
Video database.
Video file uploads and real-time streams AM access module is connect with video source.Video source is provided for assisting merit to analyze
Video resource；Video source includes video file and video flowing；Video file includes the file of the formats such as AVI, MPEG4, WMV, depending on
Frequency stream includes the live video stream of the offers such as IP camera/DVR/NVR.Video source is as the merit analysis based on security protection video
Information source, is uploaded by video file and real-time streams AM access module is imported into system and is analyzed and handled.Video file
It uploads and real-time streams AM access module stores the video file uploaded in video source to original video data library, and in video source
The live video stream of upload is saved in original video after changing into the MPEG4 video format file of standard by video format coding again
Database.
Target detection and object tracking module are for effectively extracting the motion profile of target and object in video resource
Out.In this module by specific target detection and object tracking algorithm in original video data library video file or
Person analyzes and handles from the live video stream of video source, to export moving target object (abbreviation target object)
The relevant informations such as size, shape, color, the picture snapshot of target object, motion profile of target object etc., and it is saved into picture
Snapshot and video frequency abstract index database.Finally, establishing being associated between picture snapshot and video frequency abstract index database and original video.
Target detection and object tracking algorithm are multi-target detection and track algorithm.Multi-target detection and track algorithm include
Based on multifractal spectra (Multi-Fractal Spectrum, MFS) and Robust Principal Component Analysis (Robust Principal
Component Analysis, RPCA), it realizes in the analytic process to video source, while multiple target objects, target pair
As respective motion profile analyzes processing simultaneously.In the present embodiment, on the basis of multi-target detection and track algorithm, by matching
It sets and is flexibly implanted into all kinds of intelligent video analysis algorithms, such as: intrusion detection, classification and Detection, color detection, crowd massing detection are calculated
The trigger condition of video analysis algorithm is arranged in method, when the triggering of the trigger condition of video analysis algorithm, drives merit assistant analysis
Process action.Such as: intrusion detection is to be drawn by determining regional assignment boundary in original video once there is object to swarm into
Send out intrusion detection event；Classification and Detection be according to the characteristic informations such as size, the shape of target object carry out classification and Detection, as people,
Vehicle classification；Color detection is according to color to specific objective object implementatio8 color detection；Crowd massing is detected as true according to video
The number of " human body " target occurred in region is determined to determine whether to have occurred " crowd massing " to predict the cluster event that happens suddenly.
Video analysis algoritic module executes all kinds of video analysis algorithms configured for various.Video analysis algoritic module
Execution depend on target detection and object tracking module, it may be assumed that in target detection and the implementation procedure of object tracking module lead to
It crosses the execution of configuration condition triggering associated video parser or is triggered again after target detection and object tracking module execute
The execution of associated video parser.Video analysis algorithm by the merit management on upper layer and auxiliary video figure detect module configure and
It calls, such as: it is directed to some merit associated videos, the unlatching of the video analysis algorithms such as configuration intrusion detection, crowd massing detection
Or the off option.The implementing result of video analysis algorithm, can be by as early warning response reflection to merit management and auxiliary video
Figure detects module, to support merit management and auxiliary video figure to detect.
Video summary generation module is made a summary for synthetic video as needed.Video summary generation module, which receives, comes from merit
Management and auxiliary video figure detect the configuring request of module, and the processing output according to target detection and object tracking module is (including defeated
The attribute information of target object out, the picture snapshot of target object and motion profile of target object etc.), qualified
Target object forms summarized radio (also known as concentration video) by certain video frequency abstract generating algorithm, next, abstract
Video is saved in picture snapshot and video frequency abstract index database, and establishes merit, original video, target object, summarized radio, mesh
The incidence relation between the picture snapshot of object is marked, in order to operations such as subsequent retrievals.
Merit management and auxiliary video figure detect module for carrying out integrated management to case and to case assistant analysis.Merit
Management and auxiliary video figure detect common management function that module includes merit (the common management function of merit include establish case,
Modification case deletes case, inquiry case, uploads case attachment etc.) and auxiliary figure based on video analysis detect function.It is assisting
Figure is detectd in function, by carrying out video analysis, video frequency abstract generation etc. to video source related with merit, obtains these videos
Video content and structural information, then by the association established between merit and these video structure information, for assisting merit
Analysis.The structural information of video refers mainly to moving target object included in video in the opposite position in frame between interframe
It sets, is formed with " structure " that constitutes in spatial sense；Such as: target object included in video and they size, shape,
The attribute informations such as color, classification, the structure that these objects are constituted in video frame with the relative position of interframe.
Human-computer interaction module interacts operation for realizing merit analyst and this system, can mention to merit analyst
For it is a kind of simply, conveniently, effective man-machine interactive interface, to promote the integrated pipe of the user experience of merit analyst, merit
Reason ability and analysis efficiency mainly support a variety of interactions, inquiry and retrieval function of natural language processing (NLP) and many condition input
Energy.Human-computer interaction module includes two kinds of man-machine interactive interfaces, first is that common many condition selection input or conditional attribute input,
Such as: the target object collection for meeting above-mentioned condition is screened by the way that conditional attribute is time, place, color, size, classification etc.
It closes.Another is natural language text input.Wherein, natural language processing, i.e. NLP are supported in natural language text input.For example, case
Feelings analyst can be by inputting in input frame: " 9 points to 12 points of the morning of March 10 in 2019, on Angiostrangylus cantonensis hilllock, ejection is existing
The people for wearing red jacket ", then, the natural language processing component in the human-computer interaction module of present aspect will be above-mentioned input
Text carries out accurate understanding, and to synthesize suitable SQL statement, inquiry character closes the objective result of above-mentioned implication, and returns to merit
Analyst.
Merit manages database for storing information data related with merit.
Picture snapshot and video frequency abstract index database are for efficient storage target object, characteristic information, summarized radio, movement rail
Mark.
Original video data library is for saving original video files related with case or live video stream.
Human-computer interaction module calls merit management and auxiliary video figure to detect module after the order for receiving merit analyst；
Merit management database is associated with picture snapshot and video frequency abstract index database, picture snapshot and video frequency abstract index database and original view
Frequency database association；
Merit management and auxiliary video figure detect module and call video analysis algoritic module, video summary generation module, video
Parser module, the detection of video summary generation module invocation target and object tracking module, target detection and to image tracing mould
Block analyzes original video library, and video summary generation module, target detection and object tracking module will generate result and protect
It deposits to picture snapshot and video frequency abstract index database, early warning is back to merit management and auxiliary video figure by video analysis algoritic module
Detect module；
Merit management and auxiliary video figure detect module polls merit management database, and merit manages database for query result
It is back to merit management and auxiliary video figure detects module, merit management database simultaneously saves inquiry record；
Merit management and auxiliary video figure, which detect module, will analyze result return human-computer interaction module.
As shown in figure 3, video analysis and video frequency abstract product process, detailed step are as follows:
(1) precondition: merit analyzes personnel and assigns merit management and video analysis instruction by human-computer interaction module, this
A little instructions are converted to the video analysis for video file or live video stream by merit management and auxiliary video figure frame module
With the order of video frequency abstract production.
(2) video file or live video stream are linked into target detection and object tracking module, and the module is first original
Beginning video file is stored in original video data library, and live video stream then encodes it, synchronizes and is converted to certain view
Frequency file format (such as MPEG4) is stored in original video data library.
(3) key-frame extraction is carried out by way of uniformly extracting to video file or video flowing, such as: extraction per second
4 to 12 frames differ.
(4) prospect, background (or being background) lock out operation are carried out to key frame.
(5) motion profile, the target pair of target object, target object are extracted in invocation target detection with object tracking module
As relevant information, the picture snapshot of target object etc. of (abbreviation target object).
(6) the video frequency abstract manufacturing conditions configured according to merit management and auxiliary video figure frame module, to specific mesh
Object is marked, the time response of the motion profile of pipe is fulfilled according to target, required video frequency abstract file is generated, these videos is plucked
It wants file to be written in picture snapshot and video frequency abstract index database, it is fast further to establish video source, Moving Objects, object picture
According to, the index between motion profile, video frequency abstract file, for subsequent retrieval operations use.
(7) it is finished if this video source has been analyzed, terminates this operating process；Otherwise, continue to repeat step (3) to step
Suddenly (6), until video source analysis finishes.
In (5) step, more specifically, in invocation target detection and when object tracking module, first extract target object,
The motion profile of target object；Further, continue to extract the relevant information of target object such as: the color of target object, size,
The correlations such as picture snapshot, the associated description information of motion profile, timestamp of frame where specific picture etc. and target object
Information, and being deposited into picture snapshot and video frequency abstract index database, and establish video source, target object, target object picture
Snapshot, target object motion profile between index；Further, according to merit management and auxiliary video figure frame module institute
The condition of configuration executes certain video analysis algorithm, such as: intrusion detection, target classification, the separation of people's vehicle, crowd massing detection
Scheduling algorithm, and alert event is triggered by the way that event information is written to merit management database.
Above-mentioned steps (4) and (5) be referred to as in computer vision field the preceding background separation of video, target detection with
Tracking, Objective extraction etc., there are many concrete implementation methods, such as: background subtraction, optical flow method, Kalman filtering, mean value drift
Shifting, deep approach of learning, Graph Cuts etc..The present invention during the experiment, uses background subtraction, multifractal spectra
(Multi-Fractal Spectrum, MFS) and Robust Principal Component Analysis (Robust Principal Component
Analysis, RPCA) etc. realize.
As shown in Fig. 2, a kind of intelligent merit aided analysis method based on security protection video, using above-mentioned a kind of based on security protection
The intelligent merit Computer Aided Analysis System of video, specifically includes the following steps:
(1) precondition: video analysis as shown in Figure 3 and video frequency abstract production process has been finished and merit
Relevant video resource has completed video analysis, video frequency abstract processing, has obtained corresponding data and has exported and be stored in corresponding
In database.
(2) merit analyst (such as criminal detective) inputs general query text or multiple fixations on human-computer interaction interface
Alternative condition inquired；If it is general query text, needs to carry out semantic understanding, then call natural language processing
(NLP) algorithm carries out semantic understanding processing to the query text, and inquiry request is mapped as suitable relational database SQL language
Sentence sequence；Otherwise, the association of itself and SQL statement sequence is directly established according to inquiry alternative condition.
(3) based on the SQL statement sequence in step (2), inquiry merit manages database, obtains video relevant with merit
Source identifies (video source ID).
(4) according to video source ID, in conjunction with the querying condition of the SQL statement in step (2), retrieving image snapshot and video are plucked
Index database is wanted, to obtain the picture snapshot, video frequency abstract segment and other relevant informations of qualified suspected target；Simultaneously
These contents can graphically be shown to human-computer interaction interface, and the auxiliary figure that altar table mutual affection analysis personnel do next step is detectd certainly
Plan.
(5) merit analyst is according to the picture snapshot of the suspected target of step (4), characteristic information, video frequency abstract segment etc.,
And original video is combined, carry out auxiliary merit analysis.
(6) merit analyst combines merit situation, and to suspicion object (or target), calibration after carrying out analysis decision is returned
Shelves work, and corresponding result is saved in merit management database and corresponds to case file.
(7) merit analysis personnel continue auxiliary figure if necessary and detect operation, are further continued for repeating the above steps (2) to step
Suddenly (6), if it is determined that terminating, then process terminates.
Above-described embodiment is to invent preferable embodiment, but embodiments of the present invention are not by the limit of above-described embodiment
System, other any changes, modifications, substitutions, combinations, simplifications made without departing from the spirit and principles of the present invention,
It should be equivalent substitute mode, be included within the scope of the present invention.
Claims (10)
1. a kind of intelligent merit Computer Aided Analysis System based on security protection video, it is characterised in that: including the view being connect with video source
Frequency file uploads and real-time streams AM access module, target detection and object tracking module, video analysis algoritic module, video frequency abstract life
Module, human-computer interaction module, merit management database, picture snapshot and video are detectd at module, merit management and auxiliary video figure
Abstract index database and original video data library；
The video source provides the video resource for assisting merit to analyze；
Human-computer interaction module calls merit management and auxiliary video figure to detect module after the order for receiving merit analyst；Merit
Management database is associated with picture snapshot and video frequency abstract index database, picture snapshot and video frequency abstract index database and original video number
It is associated with according to library；
Merit management and auxiliary video figure detect module and call video analysis algoritic module, video summary generation module, video analysis
Algoritic module, the detection of video summary generation module invocation target and object tracking module, target detection and object tracking module pair
Original video library is analyzed, video summary generation module, target detection and object tracking module will generate result save to
Picture snapshot and video frequency abstract index database, early warning is back to merit management to video analysis algoritic module and auxiliary video figure detects mould
Block；
Merit management and auxiliary video figure detect module polls merit management database, and merit management database returns to query result
Module is detectd to merit management and auxiliary video figure, merit management database simultaneously saves inquiry record；
Merit management and auxiliary video figure, which detect module, will analyze result return human-computer interaction module.
2. a kind of intelligent merit Computer Aided Analysis System based on security protection video according to claim 1, it is characterised in that: described
Video source includes video file and video flowing, and video file includes AVI, MPEG4, WMV format, video flowing include IP camera/
The live video stream that DVR/NVR is provided.
3. a kind of intelligent merit Computer Aided Analysis System based on security protection video according to claim 2, it is characterised in that: described
Video file uploads and real-time streams AM access module stores the video file of upload to original video data library, and real-time video
Stream is saved in original video data library by the MPEG4 video format file that video format coding changes into standard again.
4. a kind of intelligent merit Computer Aided Analysis System based on security protection video according to claim 1, it is characterised in that: described
Target detection and object tracking module divide video file or live video stream by multi-target detection and track algorithm
Analysis, processing, to export the size of target object, shape, color related information, picture snapshot, motion profile, and are saved into figure
Piece snapshot and video frequency abstract index database.
5. a kind of intelligent merit Computer Aided Analysis System based on security protection video according to claim 4, it is characterised in that: described
Multi-target detection and track algorithm include being based on multifractal spectra and Robust Principal Component Analysis, are realized in the analysis to video source
Cheng Zhong, while multiple target objects, the respective motion profile of target object are analyzed processing simultaneously.
6. a kind of intelligent merit Computer Aided Analysis System based on security protection video according to claim 5, it is characterised in that: in institute
On the basis of stating multi-target detection and track algorithm, all kinds of intelligent video analysis algorithms are flexibly implanted by configuring, it is described to be configured to
Any one or a few in intrusion detection, classification and Detection, color detection and crowd massing detection algorithm, setting video analysis are calculated
The trigger condition of method drives the process action of merit assistant analysis when the triggering of the trigger condition of video analysis algorithm.
7. a kind of intelligent merit Computer Aided Analysis System based on security protection video according to claim 1, it is characterised in that: described
Video analysis algoritic module and video summary generation module comprise the steps of:
(1) it receives merit management and video figure frame module is assisted to assign the order for video source analysis and video frequency abstract production；
(2) video file or live video stream are linked into target detection and object tracking module, and the module is first original view
Frequency file is stored in original video data library, and live video stream then encodes it, synchronizes and is converted to certain video text
Part format is stored in original video data library.
(3) file uploaded to video source carries out key-frame extraction by way of uniformly extracting；
(4) prospect is carried out to key frame, background separation operates；
(5) invocation target detection and object tracking module are to extract the motion profile of target object, target object, target object
Relevant information, the picture snapshot of target object；
(6) the video frequency abstract manufacturing conditions configured according to merit management and auxiliary video figure frame module, to specific target pair
As fulfilling the time response of the motion profile of pipe according to target, generating required video frequency abstract file, these video frequency abstracts text
Part is written in picture snapshot and video frequency abstract index database, further establishes video source, Moving Objects, object picture snapshot, fortune
Index between dynamic rail mark, video frequency abstract file, for subsequent retrieval operations use；
(7) it is finished if this video source has been analyzed, terminates this operating process；Otherwise, continue to repeat step (3) to step
(6), until video source analysis finishes.
8. a kind of intelligent merit Computer Aided Analysis System based on security protection video according to claim 1, it is characterised in that: described
Human-computer interaction module includes: (1) common many condition selection input or conditional attribute input；(2) natural language text inputs.
9. a kind of intelligent merit Computer Aided Analysis System based on security protection video according to claim 1, it is characterised in that: described
Merit management and auxiliary video figure frame module include that the common management function of merit and the auxiliary figure based on video analysis detect function；
The common management function includes establishing case, modification case, deleting case, inquiry case, upload case attachment；Described auxiliary
Figure is helped to detect in function, by carrying out video analysis to video source related with merit, video frequency abstract generates, and obtains the video of video
Content and structure information, then by establishing the association between merit and video content, structural information, for assisting case mutual affection
Analysis.
10. a kind of intelligent merit aided analysis method based on security protection video, is based on using any described one kind of claim 1-9
The intelligent merit Computer Aided Analysis System of security protection video, characterized by the following steps:
(1) merit analyst inputs general query text on human-computer interaction interface or the alternative condition of multiple fixations is looked into
It askes, if it is general query text, needs to carry out semantic understanding, then natural language processing algorithm is called to realize the query text
Semantic understanding, and inquiry request is mapped as suitable relational database SQL statement sequence；Otherwise, it is directly selected according to inquiry
Condition establishes the association of itself and SQL statement sequence；
(3) based on the SQL statement sequence in step (2), inquiry merit manages database, obtains video source mark relevant with merit
Know.
(4) according to video source mark, in conjunction with the querying condition of the SQL statement in step (2), retrieving image snapshot and video frequency abstract
Index database, to obtain picture snapshot, the video frequency abstract segment of qualified suspected target；These contents can be with figure simultaneously
Mode is shown to human-computer interaction interface, and the auxiliary figure that altar table mutual affection analysis personnel do next step detects decision；
(5) merit analyst is according to the picture snapshot of the suspected target of step (4), characteristic information, video frequency abstract segment, and combines
Original video carries out auxiliary merit analysis；
(6) merit analyst combines merit situation, to suspicion object, calibration, filing-up work after carrying out analysis decision, and phase
The result answered is saved in merit management database and corresponds to case file；
(7) merit analysis personnel continue auxiliary figure if necessary and detect operation, are further continued for repeating the above steps (2) to step
(6), if it is determined that terminating, then process terminates.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910449893.0A CN110213651A (en) | 2019-05-28 | 2019-05-28 | A kind of intelligent merit Computer Aided Analysis System and method based on security protection video |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910449893.0A CN110213651A (en) | 2019-05-28 | 2019-05-28 | A kind of intelligent merit Computer Aided Analysis System and method based on security protection video |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110213651A true CN110213651A (en) | 2019-09-06 |
Family
ID=67788933
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910449893.0A Pending CN110213651A (en) | 2019-05-28 | 2019-05-28 | A kind of intelligent merit Computer Aided Analysis System and method based on security protection video |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110213651A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111026909A (en) * | 2019-11-30 | 2020-04-17 | 杭州亿圣信息技术有限公司 | Method for supplementing public security investigation data set based on self-timer tremble view |
| CN111611522A (en) * | 2020-06-02 | 2020-09-01 | 九江学院 | Big data processing method for teaching materials of teacher education and training |
| CN113255477A (en) * | 2021-05-08 | 2021-08-13 | 深圳市安软科技股份有限公司 | Comprehensive management system and method for pedestrian video images |
| CN114915768A (en) * | 2022-07-14 | 2022-08-16 | 中通服建设有限公司 | Intelligent building security system based on passive optical network |
| CN117635402A (en) * | 2024-01-25 | 2024-03-01 | 中国人民解放军国防科技大学 | Intelligent streaming system, intelligent streaming method, intelligent streaming computer device and intelligent streaming storage medium |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101655860A (en) * | 2009-09-16 | 2010-02-24 | 北京中盾安全技术开发公司 | Video aided analytical method for criminal investigation and case detection and system |
| CN103794234A (en) * | 2012-10-30 | 2014-05-14 | 北京航天长峰科技工业集团有限公司 | Massive video-based event trace quick search platform |
| CN104021198A (en) * | 2014-06-16 | 2014-09-03 | 北京理工大学 | Relational database information retrieval method and device based on ontology semantic index |
| CN105224535A (en) * | 2014-05-29 | 2016-01-06 | 浙江航天长峰科技发展有限公司 | Based on the concern target quick reference system of massive video |
| CN107358558A (en) * | 2017-06-08 | 2017-11-17 | 上海市高级人民法院 | Criminal case intelligently handle a case method by auxiliary, system and has its storage medium and terminal device |
| CN107885786A (en) * | 2017-10-17 | 2018-04-06 | 东华大学 | Towards the Natural Language Query Interface implementation method of big data |
| CN107944371A (en) * | 2017-11-17 | 2018-04-20 | 辽宁公安司法管理干部学院 | A kind of public security video monitoring image processing method based on data mining |
| CN108052882A (en) * | 2017-11-30 | 2018-05-18 | 广东云储物联视界科技有限公司 | A kind of operating method of intelligent safety defense monitoring system |
| CN108921739A (en) * | 2018-08-06 | 2018-11-30 | 四川工商学院 | A kind of legislation intellectualized analysis platform based on big data |
-
2019
- 2019-05-28 CN CN201910449893.0A patent/CN110213651A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101655860A (en) * | 2009-09-16 | 2010-02-24 | 北京中盾安全技术开发公司 | Video aided analytical method for criminal investigation and case detection and system |
| CN103794234A (en) * | 2012-10-30 | 2014-05-14 | 北京航天长峰科技工业集团有限公司 | Massive video-based event trace quick search platform |
| CN105224535A (en) * | 2014-05-29 | 2016-01-06 | 浙江航天长峰科技发展有限公司 | Based on the concern target quick reference system of massive video |
| CN104021198A (en) * | 2014-06-16 | 2014-09-03 | 北京理工大学 | Relational database information retrieval method and device based on ontology semantic index |
| CN107358558A (en) * | 2017-06-08 | 2017-11-17 | 上海市高级人民法院 | Criminal case intelligently handle a case method by auxiliary, system and has its storage medium and terminal device |
| CN107885786A (en) * | 2017-10-17 | 2018-04-06 | 东华大学 | Towards the Natural Language Query Interface implementation method of big data |
| CN107944371A (en) * | 2017-11-17 | 2018-04-20 | 辽宁公安司法管理干部学院 | A kind of public security video monitoring image processing method based on data mining |
| CN108052882A (en) * | 2017-11-30 | 2018-05-18 | 广东云储物联视界科技有限公司 | A kind of operating method of intelligent safety defense monitoring system |
| CN108921739A (en) * | 2018-08-06 | 2018-11-30 | 四川工商学院 | A kind of legislation intellectualized analysis platform based on big data |
Non-Patent Citations (1)
| Title |
|---|
| 林龙新: ""基于对象的视频摘要算法的实现与加速"", 《华南理工大学学报》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111026909A (en) * | 2019-11-30 | 2020-04-17 | 杭州亿圣信息技术有限公司 | Method for supplementing public security investigation data set based on self-timer tremble view |
| CN111026909B (en) * | 2019-11-30 | 2023-04-28 | 杭州亿圣信息技术有限公司 | Method for supplementing public security investigation data set based on self-timer tremble sound view |
| CN111611522A (en) * | 2020-06-02 | 2020-09-01 | 九江学院 | Big data processing method for teaching materials of teacher education and training |
| CN111611522B (en) * | 2020-06-02 | 2022-07-12 | 九江学院 | Big data processing method for teaching materials of teacher education and training |
| CN113255477A (en) * | 2021-05-08 | 2021-08-13 | 深圳市安软科技股份有限公司 | Comprehensive management system and method for pedestrian video images |
| CN114915768A (en) * | 2022-07-14 | 2022-08-16 | 中通服建设有限公司 | Intelligent building security system based on passive optical network |
| CN117635402A (en) * | 2024-01-25 | 2024-03-01 | 中国人民解放军国防科技大学 | Intelligent streaming system, intelligent streaming method, intelligent streaming computer device and intelligent streaming storage medium |
| CN117635402B (en) * | 2024-01-25 | 2024-05-17 | 中国人民解放军国防科技大学 | Intelligent streaming system, intelligent streaming method, intelligent streaming computer device and intelligent streaming storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110213651A (en) | A kind of intelligent merit Computer Aided Analysis System and method based on security protection video | |
| Zaiane et al. | Mining multimedia data. | |
| CN101778260B (en) | Method and system for monitoring and managing videos on basis of structured description | |
| Fang et al. | Time series data imputation: A survey on deep learning approaches | |
| CN108052882A (en) | A kind of operating method of intelligent safety defense monitoring system | |
| CN105100718B (en) | A kind of intelligent video analysis method based on video frequency abstract | |
| CN106407278A (en) | Architecture design system of big data platform | |
| CN105809606A (en) | Big data application platform system | |
| CN106649801A (en) | Time-space relationship analysis system based on multi-source internet-of-things position awareness | |
| CN109460922A (en) | A kind of Internet public opinion analysis and aid decision-making system with power industry feature | |
| WO2017000880A1 (en) | Information processing method, information processing device and video monitoring system | |
| CN106557760A (en) | Monitoring system is filtered in a kind of image frame retrieval based on video identification technology | |
| CN106649809A (en) | Internet of Things analysis method | |
| CN103279481B (en) | Intelligent Skynet intelligence image investigation system | |
| CN113779169A (en) | Self-enhancement method of space-time data flow model | |
| CN106777279A (en) | A kind of time-space relationship analysis system | |
| CN105224535A (en) | Based on the concern target quick reference system of massive video | |
| Anwar et al. | Mining anomalous events against frequent sequences in surveillance videos from commercial environments | |
| Saleh et al. | Crime data analysis in Python using K-means clustering | |
| Chen | [Retracted] Semantic Analysis of Multimodal Sports Video Based on the Support Vector Machine and Mobile Edge Computing | |
| Renushe et al. | Data mining practices for effective investigation of crime | |
| CN107562909A (en) | A kind of big data analysis system and its analysis method for merging search and calculating | |
| Babiyola et al. | A hybrid learning frame work for recognition abnormal events intended from surveillance videos | |
| Guler et al. | A video event detection and mining framework | |
| Ram et al. | An overview on multimedia data mining and its relevance today |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20190906 |
|
| WD01 | Invention patent application deemed withdrawn after publication |