CN110119730A - A kind of monitor video processing method, system, terminal and storage medium - Google Patents
A kind of monitor video processing method, system, terminal and storage medium Download PDFInfo
- Publication number
- CN110119730A CN110119730A CN201910474481.2A CN201910474481A CN110119730A CN 110119730 A CN110119730 A CN 110119730A CN 201910474481 A CN201910474481 A CN 201910474481A CN 110119730 A CN110119730 A CN 110119730A
- Authority
- CN
- China
- Prior art keywords
- image
- video
- characteristic image
- key area
- national boundary
- 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
- 238000003672 processing method Methods 0.000 title claims abstract description 13
- 238000012544 monitoring process Methods 0.000 claims abstract description 30
- 238000000605 extraction Methods 0.000 claims abstract description 19
- 238000000034 method Methods 0.000 claims description 24
- 230000033001 locomotion Effects 0.000 claims description 19
- 230000015654 memory Effects 0.000 claims description 19
- 238000012545 processing Methods 0.000 claims description 11
- 239000000284 extract Substances 0.000 claims description 5
- 238000004590 computer program Methods 0.000 claims description 3
- 230000000977 initiatory effect Effects 0.000 claims description 3
- 238000004458 analytical method Methods 0.000 abstract description 4
- 230000005856 abnormality Effects 0.000 abstract description 2
- 238000012549 training Methods 0.000 description 14
- 230000006870 function Effects 0.000 description 5
- 238000004891 communication Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 230000000007 visual effect Effects 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000008878 coupling Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000005538 encapsulation Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Classifications
-
- 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
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B13/00—Burglar, theft or intruder alarms
- G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
- G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
- G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
- G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
- G08B13/19602—Image analysis to detect motion of the intruder, e.g. by frame subtraction
- G08B13/19613—Recognition of a predetermined image pattern or behaviour pattern indicating theft or intrusion
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
-
- 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
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- 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
- G06T2207/30196—Human being; Person
-
- 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)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Human Computer Interaction (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides a kind of monitor video processing method, system, terminal and storage medium, comprising: divides region to the video of acquisition and sets key area；Characteristic image extraction is carried out to the video；The characteristic image in key area is identified using image recognition model；Warning information is issued according to the recognition result of image recognition model output.For the present invention while reducing the workload of analysis video data, the monitoring information that can note abnormalities in time simultaneously can judge automatically whether someone passes through national boundary, and initiate alarm immediately after judging that someone passes through national boundary, play the role of timely early warning.
Description
Technical field
The invention belongs to technical field of video monitoring, and in particular to a kind of monitor video processing method, system, terminal and deposit
Storage media.
Background technique
Traditional artificial security protection video monitoring system can be realized real-time display monitored picture and video record saves, and have
Stronger safe precaution ability.But as monitoring area increase and monitor duration increase will cause under artificial quality monitoring
Drop, main cause are as follows:
The monitor video picture that cannot continue.Since national boundary is in meagrely-populated place, or it is completely in separate
The case where place of crowd, abnormal monitoring information occurs in video monitoring picture, is seldom, than attempting to wear if any nonnative personnel or animal
More national boundary.However, to ensure that the safety of national boundary, it is necessary to which someone watches monitoring screen attentively, this necessarily laborious but effect
The different mode set.Because people necessarily has carelessness when monitoring video pictures, but do not find in time once different
Normal monitoring information, it is possible to lead to unexpected consequence.
The data of magnanimity are analyzed.Since video monitoring is constantly in working condition, then necessarily will lead to monitoring data amount
Be continuously increased.Know what some time somewhere has occurred if necessary, or needs to count some necessity
Information, at this moment just need manually go analysis magnanimity video data, this will necessarily spend a large amount of manpower.
It can not timely early warning.Since traditional security protection video monitoring system is difficult to accomplish that continuous videos picture monitors, this must
It so will cause the omission of exception information, and some exception informations may result in serious consequence.If will increase without timely early warning
Add the probability for leading to serious consequence, or even if frontier officer has found exception information in time and is made that timely processing, but this
Kind response is typically all that could occur after event occurs, and can not play forewarning function.
Summary of the invention
For the above-mentioned deficiency of the prior art, the present invention provides a kind of monitor video processing method, system, terminal and storage
Medium, to solve the above technical problems.
In a first aspect, the present invention provides a kind of monitor video processing method, comprising:
Region is divided to the video of acquisition and sets key area, is specifically included: needing to be arranged away from national boundary according to monitoring
Distance range grade；Region is divided to the position in image away from national boundary distance range grade according to described；National boundary will be closed on
Region be set as key area；
Characteristic image extraction is carried out to the video, is specifically included: described in setting picture frame interception time interval and interception
The picture frame of video；Feature extraction is carried out to adjacent image frame using SIFT algorithm and judges whether adjacent image frame is consistent: being,
Then give up the preceding picture frame of cutting time point；It is no, then it saves difference adjacent image frame and extracts in picture frame key area
Characteristic image；
The characteristic image in key area is identified using image recognition model, is specifically included: being utilized
TensorFlow creates the image recognition model that can identify portrait；Using in described image identification model identification key area
Whether characteristic image is portrait and exports recognition result；
Warning information is issued according to the recognition result of image recognition model output, is specifically included: by adjacent image interframe
The higher characteristic image of the degree of association is connected in the position of picture frame obtains the motion profile of characteristic image；If the motion profile is
It is no to pass through national boundary, then early warning is initiated immediately；It intercepts and saves the video-frequency band comprising the characteristic image.
Second aspect, the present invention provide a kind of monitor video processing system, comprising:
Area division unit is configured to divide region to the video of acquisition and sets key area, comprising: grade setting
Module is configured to need to be arranged away from national boundary distance range grade according to monitoring；Area generation module is configured to according to institute
It states and region is divided to the position in image away from national boundary distance range grade；Emphasis setting module is configured to that border will be closed on
The region of line is set as key area；
Feature extraction unit is configured to carry out characteristic image extraction to the video；
Feature identification unit is configured to identify the characteristic image in key area using image recognition model；
Early warning issue unit is configured to issue warning information according to the recognition result that image recognition model exports, comprising:
Track generation module is configured in the position of picture frame mutually get the higher characteristic image of the degree of association of adjacent image interframe continuously
To the motion profile of characteristic image；Early warning initiation module, if being configured to whether the motion profile passes through national boundary, immediately
Initiate early warning；Video intercepting module is configured to intercept and save the video-frequency band comprising the characteristic image.
The third aspect provides a kind of terminal, comprising:
Processor, memory, wherein
The memory is used to store computer program,
The processor from memory for calling and running the computer program, so that terminal executes above-mentioned terminal
Method.
Fourth aspect provides a kind of computer storage medium, instruction is stored in the computer readable storage medium,
When run on a computer, so that computer executes method described in above-mentioned various aspects.
The beneficial effects of the present invention are,
Monitor video processing method, system, terminal and storage medium provided by the invention, pass through the video information to acquisition
Region division is carried out, the key monitoring to national boundary key area is realized, reduces server operation pressure, while passing through area
Domain, which divides, assists the subsequent judgement that national boundary whether is passed through to motion profile.Furthermore by video information and feature extraction and building
It is vertical to identify that the image recognition model of portrait handles video information, and then after capable of there is portrait in video information
Start immediately and portrait is locked and is tracked, this process is all that system is realized automatically.Therefore the present invention is reducing analysis view
While the workload of frequency evidence, the monitoring information that can note abnormalities in time simultaneously can judge automatically whether someone passes through border
Line, and alarm is initiated immediately after judging that someone passes through national boundary, play the role of timely early warning.
In addition, design principle of the present invention is reliable, structure is simple, has very extensive application prospect.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, for those of ordinary skill in the art
Speech, without creative efforts, is also possible to obtain other drawings based on these drawings.
Fig. 1 is the schematic flow chart of the method for one embodiment of the invention.
Fig. 2 is the schematic block diagram of the system of one embodiment of the invention.
Fig. 3 is a kind of structural schematic diagram of terminal provided in an embodiment of the present invention.
Specific embodiment
Technical solution in order to enable those skilled in the art to better understand the present invention, below in conjunction with of the invention real
The attached drawing in example is applied, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described implementation
Example is only a part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, this field is common
Technical staff's every other embodiment obtained without making creative work, all should belong to protection of the present invention
Range.
The Key Term occurred in the present invention is explained below.
Fig. 1 is the schematic flow chart of the method for one embodiment of the invention.Wherein, Fig. 1 executing subject can be one kind
Monitor video processing system.
As shown in Figure 1, this method 100 includes:
Step 110, region is divided to the video of acquisition and sets key area；
Step 120, characteristic image extraction is carried out to the video；
Step 130, the characteristic image in key area is identified using image recognition model；
Step 140, warning information is issued according to the recognition result of image recognition model output.
Optionally, as one embodiment of the invention, the video of described pair of acquisition divides region and sets key area, wraps
It includes:
It needs to be arranged away from national boundary distance range grade according to monitoring；
Region is divided to the position in image away from national boundary distance range grade according to described；
The region for closing on national boundary is set as key area.
It is optionally, described that characteristic image extraction is carried out to video as one embodiment of the invention, comprising:
Setting picture frame interception time interval and the picture frame for intercepting the video；
Feature extraction is carried out to adjacent image frame using SIFT algorithm and judges whether adjacent image frame is consistent:
It is then to give up the preceding picture frame of cutting time point；
It is no, then it saves difference adjacent image frame and extracts the characteristic image in picture frame key area.
Optionally, as one embodiment of the invention, it is described using image recognition model to the characteristic pattern in key area
As being identified, comprising:
The image recognition model of portrait can be identified using TensorFlow creation；
Whether it is portrait and exports recognition result using the characteristic image in described image identification model identification key area.
Optionally, as one embodiment of the invention, the recognition result according to the output of image recognition model issues pre-
Alert information, comprising:
The higher characteristic image of the degree of association of adjacent image interframe is connected in the position of picture frame and obtains characteristic image
Motion profile；
If whether the motion profile passes through national boundary, early warning is initiated immediately；
It intercepts and saves the video-frequency band comprising the characteristic image.
In order to facilitate the understanding of the present invention, below with the principle of monitor video processing method of the present invention, in conjunction with the embodiments
In to the process that is handled of video information of monitoring device acquisition, monitor video processing method provided by the invention is done into one
The description of step.
Specifically, monitor video processing method includes: by taking the video information for handling the acquisition of a monitor camera device as an example
S1, region is divided to the video of acquisition and sets key area.
It is divided according to the visual field that photographic device acquires video.It is arranged away from national boundary firstly the need of operator apart from model
Enclose grade.One section of national boundary of covering is allowed in the visual field of photographic device.In the present embodiment, it is arranged and the visual field is divided into four
The region division being located at left and right sides of national boundary within 50m at a distance from region, with national boundary is to close on the area Liang Ge of national boundary
The region division being located at other than 50m at left and right sides of national boundary at a distance from domain, with national boundary is two fringe regions.In the present invention
Other embodiments in also can be set it is multiple carry out more detailed region division away from national boundary distance range grade, or
Carrying out region division according to particular geographic area feature in the visual field (can such as divide the region of key monitoring and without the area of monitoring
Domain).It by the region division being located within the 50m at left and right sides of national boundary at a distance from national boundary is to close on border in the present embodiment
Two regions of line are set as key area.
S2, characteristic image extraction is carried out to the video.
In the present embodiment, in the region that national boundary needs to monitor setting photographic device (monitoring camera), monitoring camera
The video information of shooting is sent to long-range processing end (administrative center) in real time.Processing end is the load of system provided by the invention
Body executes method provided by the invention.It is as follows to the processing method of video information:
The time interval of video image frame interception is set, 5s is set as in the present embodiment, i.e., it is every by the time sequencing of video
Picture frame is opened from video intercepting one every 5s.Feature Points Matching is carried out using neighbor frame difference method to adjacent image frame, judges adjacent two
Whether frame image is consistent, when, there are when mobile object, having difference in gray scale in video between consecutive frame, seek two frames
Image grayscale absolute value of the difference, it is 0 entirely that static object shows on error image, and mobile object especially motive objects
Due to being non-zero there are grey scale change at the profile of body.Through judging, it is preceding that adjacent two field pictures give up recording time if consistent
Picture frame extracts inconsistent characteristic image using the identification of SIFT algorithm if inconsistent.Scale invariant feature converts (Scale-
Invariant feature transform or SIFT) be a kind of computer vision algorithm be used to detect and describe in image
Locality characteristic, it finds extreme point in space scale, and extracts its position, scale, rotational invariants, application range
Include object identification, the perception of robot map and navigation, image suture, the identification of 3D model foundation, gesture, image tracing and movement
It compares.SIFT algorithm can identify with lock-in feature image, so as to the subsequent tracking to characteristic image.SIFT algorithm is existing skill
Art, details are not described herein again.
S3, the characteristic image in key area is identified using image recognition model.
A large amount of character images and other image creation training sets and verifying collection are acquired, using TensorFlow on training set
Training deep learning model, and the accuracy of the model at training is verified in verifying collection.Get out training set and verifying
After collection, model is finely tuned using TensorFlow Slim, is the image classification kit that Google company announces, it is not only
Some convenient interfaces are defined, common network structure and pre-training model on many ImageNet data sets are additionally provided.
If necessary to use Slim to finely tune model, the source code of downloading Slim is first had to.In slim/datasets, define all
The database that can be used, in order to use the tfrecord data created in Section 3.2 to be trained, it is necessary to
New database is defined in datasets.After having defined data set, a satellite mesh is created again under slim file
Record, in this catalogue, completes several last preparations: creating a data catalogue, and by 3.2 sections ready 5
A training data duplication for having converted format is entered.An empty train_dir catalogue is created, for saving in training process
Log and model.A pretrained catalogue is created, is found under Inception V3 model in the GitHub page of slim
Set address after downloading and decompressing, can obtain an inception_v3.ckpt file, this document is copied to pretrained
Under catalogue.
Under slim file, operation starts training pattern to issue orders:
Trainable_scopes=InceptionV3/Logits, InceptionV3/AuxLogits: it explains first
The effect of trainable_scope, because it is extremely important.Trainable_scopes is defined finely tunes variable in a model
Range.Here setting indicates that, only to InceptionV3/Logits, two variables of InceptionV3/AuxLogits carry out micro-
It adjusts, other variables are all motionless.InceptionV3/Logits, InceptionV3/AuxLogits are equivalent in first segment
Middle said fc8, they are Inception V3 " end layers ".It, will be to mould if not setting trainable_scopes
All parameters are trained in type.
Train_dir=satellite/train_dir: show to save under satellite/train_dir catalogue
Log and checkpoint.
Dataset_name=satellite ,-dataset_split_name=train: the data set of training is specified.
The new dataset defined in 3.2 sections is exactly performance use herein.
Dataset_dir=satellite/data: the position that specified training dataset saves.
Model_name=inception_v3: the model name used.
Checkpoint_path=satellite/pretrained/inception_v3.ckpt: pre-training model
Save location.
Checkpoint_exclude_scopes=InceptionV3/Logits, InceptionV3/AuxLogi ts:
When restoring pre-training model, do not restore this two layers.It as before stated, is for this two layers the end layer of InceptionV3 model,
1000 classes and current data set for corresponding to ImageNet data set are not inconsistent, therefore not go to restore it.
- max_number_of_steps 100000: maximum execution step number.
- batch size=32: the batch quantity that every step uses.
- learning rate=0.001: learning rate.
- learning_rate_decay_type=fixed: whether learning rate declines automatically, is used herein as fixed
Habit rate.
- save interval secs=300: every 300s, program can be saved in "current" model in train dir.
It is herein exactly catalogue satellite/train dir.
- save_summaries_secs=2: every 2s, log will be written in train_dir.It can use
TensorBoard checks the log.Here for observation is facilitated, the time interval of setting is more, when hands-on, for performance
Consider, longer time interval can be set.
- log_every_n_steps=10: every 10 steps, training information will be got in screen happiness.
- optimizer=rmsprop: selected optimizer is indicated.
- weight_decay=0.00004: selected weight_decay value.I.e. in model institute's high parameter it is secondary just
Then change hyper parameter.
Multiple models are trained after executing mentioned order, need to be existed with eval_image classifier.py program at this time
It verifies and the accuracy rate of multiple models is verified under the support of collection, to obtain the highest model of accuracy rate, i.e., we need
The image recognition model that can identify portrait.The step S2 characteristic image obtained is inputted into the image recognition model, judges spy
It levies whether image is people, is transferred to step S4 if for people and enters the early warning stage, if not people sounds all clear.
S4, warning information is issued according to the recognition result of image recognition model output.
Step S3 identify characteristic image behave after, using SIFT algorithm to all picture frames after there is characteristic image into
The identification of row characteristic image, realizes the tracking to target.Then all picture frames below are sequentially arranged, by characteristic image
It is connected in chronological order in the position of different images frame, generates the motion profile of target.Judge generate motion profile whether with
National boundary has intersection point, if having intersection point immediately initiate alarm (can be in display screen display reminding information, can also be by connecting with system
The player connect initiates audio alarm).In addition, being intercepted corresponding with the picture frame generated while system initiates alarm
Video-frequency band, and by the video push of interception to administrative staff (pushing to long-range processing end display screen).If motion profile and border
Line does not have intersection point then to continue tracking feature image, at this time in system there are chronological set of frames, first in set
Picture frame is that normal not characteristic image, the second picture frame start characteristic image occur, is newly generated when tracking
Picture frame is identical as first picture frame, and after characteristic image, and the motion profile and national boundary that generate do not have intersection point, then solve
Except alarm.
In other embodiments of the invention, national boundary is passed through in order to prevent someone in advance, it can be with national boundary or so two
The line of demarcation in two adjacent regions of side judges whether target trajectory with line of demarcation has intersection point as Alert Standard, into
And early warning is carried out, it in this way can further crime prevention.
In addition above-described embodiment be by taking the monitor video of a camera as an example, in other embodiments of the invention,
To the same monitoring area of national boundary, camera can be installed from multiple angles, acquire the multi-angle monitoring of same monitoring area
The monitor video of video, mostly each angle shot uses the above method to be handled, while generating the three of same characteristic image
Information is tieed up, the accuracy rate to Identification of Images in monitor video is improved.
If Fig. 2 shows, which includes:
Area division unit 210, the area division unit 210 are used to divide region to the video of acquisition and set emphasis
Region；
Feature extraction unit 220, the feature extraction unit 220 are used to carry out characteristic image extraction to the video；
Feature identification unit 230, the feature identification unit 230 are used for using image recognition model in key area
Characteristic image is identified；
Early warning issue unit 240, the early warning issue unit 240 are used for the recognition result exported according to image recognition model
Issue warning information.
Optionally, as one embodiment of the invention, the area division unit includes:
Grade setup module is configured to need to be arranged away from national boundary distance range grade according to monitoring；
Area generation module is configured to according to described away from national boundary distance range grade to the position dividing regions in image
Domain；
Emphasis setting module is configured to the region for closing on national boundary being set as key area.
Optionally, as one embodiment of the invention, the early warning issue unit includes:
Track generation module is configured to the higher characteristic image of the degree of association of adjacent image interframe in the position of picture frame
It sets to be connected and obtains the motion profile of characteristic image；
Early warning initiation module initiates early warning if being configured to whether the motion profile passes through national boundary immediately；
Video intercepting module is configured to intercept and save the video-frequency band comprising the characteristic image.
Fig. 3 is a kind of structural schematic diagram of terminal system 300 provided in an embodiment of the present invention, which can be with
For executing monitor video processing method provided in an embodiment of the present invention.
Wherein, which may include: processor 310, memory 320 and communication unit 330.These components
It is communicated by one or more bus, it will be understood by those skilled in the art that the structure of server shown in figure is not
Limitation of the invention is constituted, it is also possible to hub-and-spoke configuration either busbar network, can also include more than illustrating
Or less component, perhaps combine certain components or different component layouts.
Wherein, which can be used for executing instruction for storage processor 310, and memory 320 can be by any class
The volatibility or non-volatile memories terminal or their combination of type are realized, such as static random access memory (SRAM), electricity
Erasable Programmable Read Only Memory EPROM (EEPROM), Erasable Programmable Read Only Memory EPROM (EPROM), programmable read only memory
(PROM), read-only memory (ROM), magnetic memory, flash memory, disk or CD.When executing instruction in memory 320
When being executed by processor 310, so that terminal 300 some or all of is able to carry out in following above method embodiment step.
Processor 310 is the control centre for storing terminal, utilizes each of various interfaces and the entire electric terminal of connection
A part by running or execute the software program and/or module that are stored in memory 320, and calls and is stored in storage
Data in device, to execute the various functions and/or processing data of electric terminal.The processor can be by integrated circuit
(Integrated Circuit, abbreviation IC) composition, such as the IC that can be encapsulated by single are formed, can also be by more of connection
The encapsulation IC of identical function or different function and form.For example, processor 310 can only include central processing unit
(Central Processing Unit, abbreviation CPU).In embodiments of the present invention, CPU can be single operation core, can also
To include multioperation core.
Communication unit 330, for establishing communication channel, so that the storage terminal be allow to be led to other terminals
Letter.It receives the user data of other terminals transmission or sends user data to other terminals.
The present invention also provides a kind of computer storage mediums, wherein the computer storage medium can be stored with program, the journey
Sequence may include step some or all of in each embodiment provided by the invention when executing.The storage medium can for magnetic disk,
CD, read-only memory (English: read-only memory, referred to as: ROM) or random access memory (English:
Random access memory, referred to as: RAM) etc..
Therefore, the present invention is realized by carrying out region division to the video information of acquisition to national boundary key area
Key monitoring, reduces server operation pressure, while assisting subsequent whether passing through border to motion profile by region division
The judgement of line.Furthermore by video information and feature extraction and establishing and can identify that the image recognition model of portrait believes video
Breath is handled, and then is started immediately to portrait locking and tracked after capable of occurring portrait in video information, this process is all
It is that system is realized automatically.Therefore the present invention can have found different in time while reducing the workload of analysis video data
Normal monitoring information simultaneously can judge automatically whether someone passes through national boundary, and initiate immediately after judging that someone passes through national boundary
Alarm, plays the role of timely early warning, the attainable technical effect of the present embodiment institute may refer to it is described above, herein
It repeats no more.
It is required that those skilled in the art can be understood that the technology in the embodiment of the present invention can add by software
The mode of general hardware platform realize.Based on this understanding, the technical solution in the embodiment of the present invention substantially or
Say that the part that contributes to existing technology can be embodied in the form of software products, which is stored in
Such as USB flash disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory in one storage medium
The various media that can store program code such as (RAM, Random Access Memory), magnetic or disk, including it is several
Instruction is used so that a terminal (can be personal computer, server or second terminal, the network terminal etc.) is held
Row all or part of the steps of the method according to each embodiment of the present invention.
Same and similar part may refer to each other between each embodiment in this specification.Implement especially for terminal
For example, since it is substantially similar to the method embodiment, so being described relatively simple, related place is referring in embodiment of the method
Explanation.
In several embodiments provided by the present invention, it should be understood that disclosed system, system and method, it can be with
It realizes by another way.For example, system embodiment described above is only schematical, for example, the unit
It divides, only a kind of logical function partition, there may be another division manner in actual implementation, such as multiple units or components
It can be combined or can be integrated into another system, or some features can be ignored or not executed.Another point, it is shown or
The mutual coupling, direct-coupling or communication connection discussed can be through some interfaces, the indirect coupling of system or unit
It closes or communicates to connect, can be electrical property, mechanical or other forms.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme
's.
It, can also be in addition, the functional units in various embodiments of the present invention may be integrated into one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.
Although by reference to attached drawing and combining the mode of preferred embodiment to the present invention have been described in detail, the present invention
It is not limited to this.Without departing from the spirit and substance of the premise in the present invention, those of ordinary skill in the art can be to the present invention
Embodiment carry out various equivalent modifications or substitutions, and these modifications or substitutions all should in covering scope of the invention/appoint
What those familiar with the art in the technical scope disclosed by the present invention, can easily think of the change or the replacement, answer
It is included within the scope of the present invention.Therefore, protection scope of the present invention is answered described is with scope of protection of the claims
It is quasi-.
Claims (10)
1. a kind of monitor video processing method characterized by comprising
Region is divided to the video of acquisition and sets key area；
Characteristic image extraction is carried out to the video；
The characteristic image in key area is identified using image recognition model；
Warning information is issued according to the recognition result of image recognition model output.
2. the method according to claim 1, wherein the video of described pair of acquisition divides region and sets emphasis area
Domain, comprising:
It needs to be arranged away from national boundary distance range grade according to monitoring；
Region is divided to the position in image away from national boundary distance range grade according to described；
The region for closing on national boundary is set as key area.
3. the method according to claim 1, wherein described carry out characteristic image extraction to video, comprising:
Setting picture frame interception time interval and the picture frame for intercepting the video；
Feature extraction is carried out to adjacent image frame using SIFT algorithm and judges whether adjacent image frame is consistent:
It is then to give up the preceding picture frame of cutting time point；
It is no, then it saves difference adjacent image frame and extracts the characteristic image in picture frame key area.
4. the method according to claim 1, wherein it is described using image recognition model to the spy in key area
Sign image is identified, comprising:
The image recognition model of portrait can be identified using TensorFlow creation；
Whether it is portrait and exports recognition result using the characteristic image in described image identification model identification key area.
5. the method according to claim 1, wherein the recognition result according to the output of image recognition model is sent out
Warning information out, comprising:
The higher characteristic image of the degree of association of adjacent image interframe is connected in the position of picture frame and obtains the movement of characteristic image
Track；
If whether the motion profile passes through national boundary, early warning is initiated immediately；
It intercepts and saves the video-frequency band comprising the characteristic image.
6. a kind of monitor video processing system characterized by comprising
Area division unit is configured to divide region to the video of acquisition and sets key area；
Feature extraction unit is configured to carry out characteristic image extraction to the video；
Feature identification unit is configured to identify the characteristic image in key area using image recognition model；
Early warning issue unit is configured to issue warning information according to the recognition result that image recognition model exports.
7. system according to claim 6, which is characterized in that the area division unit includes:
Grade setup module is configured to need to be arranged away from national boundary distance range grade according to monitoring；
Area generation module is configured to divide region to the position in image away from national boundary distance range grade according to described；
Emphasis setting module is configured to the region for closing on national boundary being set as key area.
8. system according to claim 6, which is characterized in that the early warning issue unit includes:
Track generation module, be configured to by the higher characteristic image of the degree of association of adjacent image interframe picture frame position phase
Get the motion profile of characteristic image continuously；
Early warning initiation module initiates early warning if being configured to whether the motion profile passes through national boundary immediately；
Video intercepting module is configured to intercept and save the video-frequency band comprising the characteristic image.
9. a kind of terminal characterized by comprising
Processor；
The memory executed instruction for storage processor；
Wherein, the processor is configured to perform claim requires the described in any item methods of 1-5.
10. a kind of computer readable storage medium for being stored with computer program, which is characterized in that the program is executed by processor
Shi Shixian method according to any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910474481.2A CN110119730A (en) | 2019-06-03 | 2019-06-03 | A kind of monitor video processing method, system, terminal and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910474481.2A CN110119730A (en) | 2019-06-03 | 2019-06-03 | A kind of monitor video processing method, system, terminal and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110119730A true CN110119730A (en) | 2019-08-13 |
Family
ID=67523598
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910474481.2A Pending CN110119730A (en) | 2019-06-03 | 2019-06-03 | A kind of monitor video processing method, system, terminal and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110119730A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110599737A (en) * | 2019-09-06 | 2019-12-20 | 杭州思源信息技术股份有限公司 | Social comprehensive public safety active prevention system |
| CN110942578A (en) * | 2019-11-29 | 2020-03-31 | 韦达信息技术(深圳)有限公司 | Intelligent analysis anti-theft alarm system |
| CN111601080A (en) * | 2020-05-12 | 2020-08-28 | 杭州武盛广告制作有限公司 | Video management system for community security monitoring video storage |
| CN112001892A (en) * | 2020-07-29 | 2020-11-27 | 苏州浪潮智能科技有限公司 | Server accessory supervision method, system, terminal and storage medium |
Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101060624A (en) * | 2007-05-08 | 2007-10-24 | 杭州华三通信技术有限公司 | Video data processing method and storage equipment |
| CN101123721A (en) * | 2007-09-30 | 2008-02-13 | 湖北东润科技有限公司 | An intelligent video monitoring system and its monitoring method |
| CN101295405A (en) * | 2008-06-13 | 2008-10-29 | 西北工业大学 | Portrait and vehicle recognition alarming and tracing method |
| CN201226173Y (en) * | 2008-06-27 | 2009-04-22 | 天津三星电子有限公司 | Video monitoring apparatus with automatic monitoring alarm function |
| CN101916448A (en) * | 2010-08-09 | 2010-12-15 | 云南清眸科技有限公司 | Moving object detecting method based on Bayesian frame and LBP (Local Binary Pattern) |
| CN105959637A (en) * | 2016-06-01 | 2016-09-21 | 福州大学 | Remote intelligent video monitoring system |
| CN106204646A (en) * | 2016-07-01 | 2016-12-07 | 湖南源信光电科技有限公司 | Multiple mobile object tracking based on BP neutral net |
| CN107195198A (en) * | 2017-07-31 | 2017-09-22 | 重庆酷贝科技发展有限公司 | A kind of curb parking management method and its system based on image recognition technology |
| CN107205137A (en) * | 2017-07-07 | 2017-09-26 | 广东中星电子有限公司 | One kind monitoring recognition methods and device |
| CN107818651A (en) * | 2017-10-27 | 2018-03-20 | 华润电力技术研究院有限公司 | A kind of illegal cross-border warning method and device based on video monitoring |
| CN108230364A (en) * | 2018-01-12 | 2018-06-29 | 东南大学 | A kind of foreground object motion state analysis method based on neural network |
| CN108985170A (en) * | 2018-06-15 | 2018-12-11 | 山东信通电子股份有限公司 | Transmission line of electricity hanger recognition methods based on Three image difference and deep learning |
| CN109377703A (en) * | 2018-12-06 | 2019-02-22 | 河池学院 | A kind of forest fireproofing early warning system and its method based on machine vision |
| CN109410496A (en) * | 2018-10-25 | 2019-03-01 | 北京交通大学 | Attack early warning method, apparatus and electronic equipment |
| CN109614906A (en) * | 2018-12-03 | 2019-04-12 | 北京工业大学 | A kind of security system and security alarm method based on deep learning |
-
2019
- 2019-06-03 CN CN201910474481.2A patent/CN110119730A/en active Pending
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101060624A (en) * | 2007-05-08 | 2007-10-24 | 杭州华三通信技术有限公司 | Video data processing method and storage equipment |
| CN101123721A (en) * | 2007-09-30 | 2008-02-13 | 湖北东润科技有限公司 | An intelligent video monitoring system and its monitoring method |
| CN101295405A (en) * | 2008-06-13 | 2008-10-29 | 西北工业大学 | Portrait and vehicle recognition alarming and tracing method |
| CN201226173Y (en) * | 2008-06-27 | 2009-04-22 | 天津三星电子有限公司 | Video monitoring apparatus with automatic monitoring alarm function |
| CN101916448A (en) * | 2010-08-09 | 2010-12-15 | 云南清眸科技有限公司 | Moving object detecting method based on Bayesian frame and LBP (Local Binary Pattern) |
| CN105959637A (en) * | 2016-06-01 | 2016-09-21 | 福州大学 | Remote intelligent video monitoring system |
| CN106204646A (en) * | 2016-07-01 | 2016-12-07 | 湖南源信光电科技有限公司 | Multiple mobile object tracking based on BP neutral net |
| CN107205137A (en) * | 2017-07-07 | 2017-09-26 | 广东中星电子有限公司 | One kind monitoring recognition methods and device |
| CN107195198A (en) * | 2017-07-31 | 2017-09-22 | 重庆酷贝科技发展有限公司 | A kind of curb parking management method and its system based on image recognition technology |
| CN107818651A (en) * | 2017-10-27 | 2018-03-20 | 华润电力技术研究院有限公司 | A kind of illegal cross-border warning method and device based on video monitoring |
| CN108230364A (en) * | 2018-01-12 | 2018-06-29 | 东南大学 | A kind of foreground object motion state analysis method based on neural network |
| CN108985170A (en) * | 2018-06-15 | 2018-12-11 | 山东信通电子股份有限公司 | Transmission line of electricity hanger recognition methods based on Three image difference and deep learning |
| CN109410496A (en) * | 2018-10-25 | 2019-03-01 | 北京交通大学 | Attack early warning method, apparatus and electronic equipment |
| CN109614906A (en) * | 2018-12-03 | 2019-04-12 | 北京工业大学 | A kind of security system and security alarm method based on deep learning |
| CN109377703A (en) * | 2018-12-06 | 2019-02-22 | 河池学院 | A kind of forest fireproofing early warning system and its method based on machine vision |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110599737A (en) * | 2019-09-06 | 2019-12-20 | 杭州思源信息技术股份有限公司 | Social comprehensive public safety active prevention system |
| CN110942578A (en) * | 2019-11-29 | 2020-03-31 | 韦达信息技术(深圳)有限公司 | Intelligent analysis anti-theft alarm system |
| CN111601080A (en) * | 2020-05-12 | 2020-08-28 | 杭州武盛广告制作有限公司 | Video management system for community security monitoring video storage |
| CN112001892A (en) * | 2020-07-29 | 2020-11-27 | 苏州浪潮智能科技有限公司 | Server accessory supervision method, system, terminal and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11210504B2 (en) | Emotion detection enabled video redaction | |
| Tripathi et al. | Convolutional neural networks for crowd behaviour analysis: a survey | |
| CN110119730A (en) | A kind of monitor video processing method, system, terminal and storage medium | |
| Kumar et al. | The p-destre: A fully annotated dataset for pedestrian detection, tracking, and short/long-term re-identification from aerial devices | |
| US20190325198A1 (en) | Surveillance and monitoring system that employs automated methods and subsystems that identify and characterize face tracks in video | |
| CN101778260B (en) | Method and system for monitoring and managing videos on basis of structured description | |
| US20210382933A1 (en) | Method and device for archive application, and storage medium | |
| WO2021139324A1 (en) | Image recognition method and apparatus, computer-readable storage medium and electronic device | |
| Kolar Rajagopal et al. | Exploring transfer learning approaches for head pose classification from multi-view surveillance images | |
| CN111627045A (en) | Multi-pedestrian online tracking method, device and equipment under single lens and storage medium | |
| Hashmi et al. | An exploratory analysis on visual counterfeits using conv-lstm hybrid architecture | |
| CN110659391A (en) | Video detection method and device | |
| US8798318B2 (en) | System and method for video episode viewing and mining | |
| CN110472516A (en) | A kind of construction method, device, equipment and the system of character image identifying system | |
| US11113838B2 (en) | Deep learning based tattoo detection system with optimized data labeling for offline and real-time processing | |
| WO2021135138A1 (en) | Target motion trajectory construction method and device, and computer storage medium | |
| US20230060211A1 (en) | System and Method for Tracking Moving Objects by Video Data | |
| KR20190088087A (en) | method of providing categorized video processing for moving objects based on AI learning using moving information of objects | |
| Nakahata et al. | Anomaly detection with a moving camera using spatio-temporal codebooks | |
| Apostolakis et al. | DARLENE–Improving situational awareness of European law enforcement agents through a combination of augmented reality and artificial intelligence solutions | |
| US20210089784A1 (en) | System and Method for Processing Video Data from Archive | |
| CN115862113A (en) | Stranger abnormity identification method, device, equipment and storage medium | |
| Loukil et al. | Design and test of smart IP-camera within reconfigurable platform | |
| CN111797175A (en) | Data storage method and device, storage medium and electronic equipment | |
| Teja et al. | Man-on-man brutality identification on video data using Haar cascade algorithm |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190813 |