CN108540707A - Recognition of face crime scene investigation device - Google Patents
Recognition of face crime scene investigation device Download PDFInfo
- Publication number
- CN108540707A CN108540707A CN201810732512.5A CN201810732512A CN108540707A CN 108540707 A CN108540707 A CN 108540707A CN 201810732512 A CN201810732512 A CN 201810732512A CN 108540707 A CN108540707 A CN 108540707A
- Authority
- CN
- China
- Prior art keywords
- image
- face
- recognition
- facial image
- module
- 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
- 238000011835 investigation Methods 0.000 title claims abstract description 37
- 230000001815 facial effect Effects 0.000 claims abstract description 90
- 238000003384 imaging method Methods 0.000 claims abstract description 18
- 238000004891 communication Methods 0.000 claims abstract description 17
- 230000005540 biological transmission Effects 0.000 claims abstract description 14
- 230000003287 optical effect Effects 0.000 claims abstract description 5
- 238000006243 chemical reaction Methods 0.000 claims abstract description 4
- 238000000034 method Methods 0.000 claims description 25
- 238000007781 pre-processing Methods 0.000 claims description 18
- 238000001514 detection method Methods 0.000 claims description 7
- 238000001914 filtration Methods 0.000 claims description 7
- 238000012216 screening Methods 0.000 claims description 6
- 238000012544 monitoring process Methods 0.000 abstract description 15
- 238000012545 processing Methods 0.000 description 13
- 238000010586 diagram Methods 0.000 description 8
- 238000012986 modification Methods 0.000 description 4
- 230000004048 modification Effects 0.000 description 4
- 238000009434 installation Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 230000006835 compression Effects 0.000 description 2
- 238000007906 compression Methods 0.000 description 2
- 238000012790 confirmation Methods 0.000 description 2
- 235000013399 edible fruits Nutrition 0.000 description 2
- 230000005611 electricity Effects 0.000 description 2
- 230000003321 amplification Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000010008 shearing Methods 0.000 description 1
- 230000008054 signal transmission Effects 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/50—Constructional details
-
- 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
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/172—Classification, e.g. identification
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/50—Constructional details
- H04N23/54—Mounting of pick-up tubes, electronic image sensors, deviation or focusing coils
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/50—Constructional details
- H04N23/55—Optical parts specially adapted for electronic image sensors; Mounting thereof
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/60—Control of cameras or camera modules
- H04N23/61—Control of cameras or camera modules based on recognised objects
- H04N23/611—Control of cameras or camera modules based on recognised objects where the recognised objects include parts of the human body
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/60—Control of cameras or camera modules
- H04N23/67—Focus control based on electronic image sensor signals
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/80—Camera processing pipelines; Components thereof
-
- 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
- H04N7/181—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast for receiving images from a plurality of remote sources
Abstract
The present invention provides a kind of recognition of face crime scene investigation device, including recognition of face camera, memory and wireless communication module, recognition of face camera includes camera body, camera body includes automatically controlled wide-angle lens, imaging sensor, controller and facial image pretreatment and captures module, the optical signal of automatically controlled wide-angle lens is converted to digital signal by imaging sensor, imaging sensor is by the digital data transmission of conversion to controller, controller connects facial image pretreatment and captures module, and controller is also respectively connected with memory and wireless communication module;The present invention can effectively reduce monitoring cost, greatly reduce the required data volume transmitted, improve transmission efficiency, the facial image of high value is provided for the backstage recognition of face server or portable equipment in a distant place.
Description
Technical field
The present invention relates to a kind of recognition of face crime scene investigation devices.
Background technology
Existing recognition of face server can accomplish the comparison of two face pictures, typically a face photographed in real time
The standard certificate photograph of image and a lane database compares, and it is whom to find out this present people, such as airport and high ferro station,
It is required that passenger takes pictures the face safety check camera sufficiently long time, then compared with the identity card of submission, confirmation is same person.
However, in more application scenarios, such as public place monitoring system and crime scene investigation device etc., it is impossible to it is required that each
The people passed by stands to a bat standard photograph before camera, then waits for server is allowed to identify within tens seconds.In monitoring system and hold
On method instrument, the image that camera obtains is intricate, fast changing, and many times face is all invisible at all.And it images
How high resolution digital video image is sent to recognition of face server by the position of head installation all far from control platform, and
How to obtain facial image is the bottleneck that monitoring system realizes recognition of face.
To solve the problems, such as high resolution digital video image transmitting, there are two types of present ways, and one is face is known
Other server is put into the place of camera installation, and the high resolution video image of camera is directly accessed server and is carried out to image
Processing and identification;Another method is that after the high-definition digital video image of camera is handled into overcompression, have by special
Line high-speed channel is sent to the control platform in a distant place, then is linked into recognition of face server and is handled and identified.
The above both methods cannot all meet the requirement of practical application.In former scheme, each camera exists
One expensive server of in-site installation is unrealistic, and the database in server is very high certificate of confidentiality requirement etc.
Information, it is easy to cause to divulge a secret, along with the content of these databases is controlled by control platform, that is to say, that these distributions
It must be connected with the Intranet of control platform in the server of each camera, this cannot meet the requirement of network security.Latter
In kind scheme, since the data volume of digital video image is very big, the data of 4K high resolution video images 12Gb about per second.
Data volume big in this way cannot be transmitted in existing telecommunication path, need to be compressed to original thousand with the compression algorithms such as H.265
/ mono- in special wired paths hereinafter, could transmit.But high resolution video image is after overcompression, image it is thin
Section is largely lost, and significantly reduces the accuracy of recognition of face, especially when there is multiple faces in image, each face
Information content compressed after can not identify.In the wearable devices such as crime scene investigation device, since transmission mode can only be wireless, and
Can only be battery powered, image transmitting problem is more serious, and the video image of camera needs a ten thousandth or less to be possible to
Transmission.And wireless transmission expense is very high, cannot meet real requirement.
Refering to what is shown in Fig. 1, Fig. 1 is the structural schematic diagram of existing face identification system, server and Image Acquisition camera
In the same place, server is filtered camera the image collected, object separation, the processing such as Face datection, people
It after face is separated, is compared with the picture of identity card, confirmation is same person.Such system structure can be used for airport and
The special safety check bayonet such as high ferro has enough spaces, there is special messenger guard because these places belong to special harbor,
Server and database are all relatively safer.But such system cannot be used in public place monitoring system, such as road, square,
The places such as market, the place for installing camera do not have place to put server.Even if there is place to put server, there is also prodigious peaces
Full hidden danger cannot meet application requirement.Especially server need with the control platform network connection in a distant place, on these networks
The data such as certificate belong to private data, and the data that external equipment accesses these Intranets can cause many security breaches, violate number
According to security regulations.
Refering to what is shown in Fig. 2, indicating the structural schematic diagram of another existing face identification system, Image Acquisition camera passes through
Private telecommunication network sends video image to the control platform in a distant place, and server is filtered camera the image collected, object
Body detaches, and the processing such as Face datection after face is separated, are compared with the picture of identity card, find out the detailed of passing people
Thin information.But high-resolution digital video image is transferred to a distant place, it is necessary to calculate by a large amount of compressions, such as H.265 etc.
Data volume is compressed to original one thousandth hereinafter, the result is that a large amount of details of image are lost, can only many times seen by method
To the fuzzy shadow, the accuracy of recognition of face is not caused to seriously affect.And in such system image delay it is very long,
Usually want several seconds, etc. control platforms when identifying face, examined people has had been moved off scene, cannot meet high request
Face recognition application.
In both the above system structure, server spends a large amount of calculation resources to be regarded for handling high-resolution digital
Frequently, 100 twentieth images can only be handled in this way, that is to say, that detect that the possibility of face in image is small
In 1%, it requires that face face camera is more than two seconds time, can just find an effective facial image.
Recognition of face is an extremely complex and time-consuming process, it is desirable that high performance server hardware and artificial intelligence are soft
Part.By existing computing capability, the digital picture of a 4K resolution ratio is handled, it is determined whether with the presence of identifiable face, greatly
It is general to need 2 seconds, if there is identifiable face exists, face is separated, is treated as with certificate photograph to be compared
To format, each facial image probably need 1 second.Later using the photo on complicated intelligent algorithm and certificate into
Row compares, it is determined whether consistent.However digital video is 60 width image per second, this results in recognition of face server can only be every
Piece image is handled in 120 width images.This processed image is entirely random inside video flowing.That is it is
Ensure to find one and can be used in the facial image people that identification compares and must keep face face camera 2 seconds or more.Then pass through
It can determine within several seconds whether face is consistent with certificate.This disclosure satisfy that the application on airport and high ferro safe examination system, can require
Each passenger's face, which images several leading second, to be identified.But in the wearable devices such as public's monitoring system and crime scene investigation device, in this way
Be not required to meet.There are many people in video image, and all ceaselessly moving, faces the chance little time of camera
It is very short.Piece image in 120 width image of recognition of face server random process, can find the chance for the face that can be identified
Very little, and many computing resources are wasted for handling nugatory image.
The above problem should be paid attention to and be solved the problems, such as during the video monitoring of mobile population.
Invention content
Solve existing in the prior art how convection current is moving the object of the present invention is to provide a kind of recognition of face crime scene investigation device
Group carries out the problem of fast and accurately high resolution video surveillance.
Technical solution of the invention is:
A kind of recognition of face crime scene investigation device, including recognition of face camera, memory and wireless communication module, recognition of face camera
Including camera body, camera body include automatically controlled wide-angle lens, imaging sensor, controller and facial image pretreatment and
Module is captured, the optical signal of automatically controlled wide-angle lens is converted to digital signal by imaging sensor, and imaging sensor is by the number of conversion
Word signal transmission is to controller, and controller connects facial image pretreatment and captures module, and controller is also respectively connected with memory
And wireless communication module.
Further, recognition of face camera further includes the electron cloud realized by automatically controlled wide-angle lens and imaging sensor
Platform, automatically controlled wide-angle lens project image onto on imaging sensor, and facial image pretreatment and candid photograph module control electronic platform
It tracks and amplifies face.
Further, controller, facial image pretreatment and candid photograph module, memory and wireless communication module are set to electricity
On the plate of road, circuit board is additionally provided with power supply and USB interface.
Further, wireless communication module is communicated by telecommunications network with control platform.
Further, automatically controlled wide-angle lens passes through high-definition digital video with the collected real-time digital video of imaging sensor
Controller accesses facial image pretreatment and captures module;Facial image pre-processes and captures module and locates in advance to real-time digital video
After reason, the image for the face that can be used for matching identification is filtered out, or track by controlling electronic platform and automatically controlled wide-angle lens
The facial image of optimum state is captured, then facial image pretreatment and captures the matching identification of can be used for that module will filter out
The image of face by image cache module cache and sends control platform to by network interface.
Further, facial image pretreatment and candid photograph module carry out image preprocessing to real-time digital video, specifically,
The foreground and background that original image is detached using differential technique is divided the middle foreground image of original image and background image
From obtaining foreground image, the boundary profile that convolution algorithm finds out object carried out to foreground image and high-pass filtering template, according to wheel
Wide continuity isolates independent object with closure, and the image of each object is normalized, then special with face
It levies template and carries out convolution algorithm, obtain subject image.
Further, facial image pretreatment and candid photograph module filter out the image for the face that can be used for matching identification, tool
Body is after carrying out image preprocessing to real-time digital video, to obtain subject image, judge whether the subject image is positive dough figurine
Face, if clarity is not less than setting value, filters out if so, seeing the clarity of facial image again and can be used for matching identification
Face image.
Further, facial image pretreatment and candid photograph module capture optimum state facial image, specifically, to real-time
After digital video carries out image preprocessing, subject image is obtained, judges whether the subject image is front face, if so, again
See the clarity of facial image, if clarity is less than setting value, control electronic platform is directed at the face location, and controls automatically controlled
Wide-angle lens amplifies and focuses on the face, then according to facial image and positive eye opening skin detection convolution as a result,
Capture the facial image of optimum state.
Further, facial image is pre-processed and is captured in module, after carrying out image preprocessing to real-time digital video, is obtained
To subject image, judge whether the subject image is front face, be not front face if it is the subject image, then deleting should
Subject image.
Further, control platform receives the facial image of transmission by telecommunications network, and access face server is to these people
Face image is identified.
The beneficial effects of the invention are as follows:
One, this kind of recognition of face crime scene investigation device can effectively reduce monitoring cost, greatly reduce the required data volume transmitted, carry
High transmission efficiency improves the face recognition accuracy rate and speed of monitoring system.This kind of recognition of face crime scene investigation device can be a distant place
Backstage recognition of face server or portable equipment provide the facial image of high value.This kind of recognition of face crime scene investigation device, can be in height
In the digital video image of resolution ratio or ultrahigh resolution, facial image is found out in real time, and sends backstage to and processing is identified.
Two, the present invention utilizes high-speed figure image processing techniques, the real-time digital video in recognition of face camera locality
Middle that people face image is separated from background image, control recognition of face camera tracks and amplifies valuable people face, when
When obtaining complete clearly people face image, captures and preserve the people face image, and by communication networks such as mobile phone wireless nets, people face figure
It is identified as being sent to distant place background server, carries out front end High-speed video images pretreatment for face identification system, screening is simultaneously
Capture clearly people face image, greatlys improve the efficiency of backstage human bioequivalence server, and filters out that account about video data total
The rubbish image of amount 99.99%, well solves communication bandwidth and image clearly needed for thecamera head to distant place background server
The contradiction of degree.
Three, in the present invention, the collected digital video of recognition of face crime scene investigation device accesses preprocessed chip, by real-time digital
After video pre-filtering, effective facial image is filtered out, or control electronic platform and the tracking of automatically controlled wide-angle lens are captured most preferably
Facial image is transferred to control platform, access face server is to these figures then these effective images by telecommunications network
As being identified.For digital video image after screening, effective face image data about only has original video data amount
A ten thousandth can effectively utilize existing telecommunication network or wireless network transmissions.And since camera completes object
Body identifies and the work such as facial image screening, and recognition of face server can be sufficiently used for operational capability face alignment identification etc.
Follow-up work, a server can handle the input picture of multiple cameras, and the fund of expensive server is greatly saved
Input.
Description of the drawings
Fig. 1 is structural schematic diagram of the existing recognition of face server in the local system of monitoring.
Fig. 2 is the structural schematic diagram of system of the existing recognition of face server in control platform.
Fig. 3 is that existing crime scene investigation device illustrates block diagram.
Fig. 4 is that recognition of face crime scene investigation device of the embodiment of the present invention illustrates block diagram.
Fig. 5 is that mobile population video monitoring system of the embodiment based on recognition of face illustrates schematic diagram.
Fig. 6 is the flow diagram of the image pre-processing method of facial image pretreatment and candid photograph module in embodiment.
Specific implementation mode
The preferred embodiment that the invention will now be described in detail with reference to the accompanying drawings.
Embodiment
A kind of recognition of face crime scene investigation device, such as Fig. 4, including recognition of face camera, memory and wireless communication module, face
Identification camera includes camera body, and camera body includes automatically controlled wide-angle lens, imaging sensor, controller and face figure
As pre-processing and capturing module, the optical signal of automatically controlled wide-angle lens is converted to digital signal, imaging sensor by imaging sensor
By the digital data transmission of conversion to controller, controller connects facial image pretreatment and captures module, and controller is also distinguished
Connect memory and wireless communication module.
This kind of recognition of face crime scene investigation device, can effectively reduce monitoring cost, greatly reduce the required data volume transmitted, carry
High transmission efficiency improves the face recognition accuracy rate and speed of monitoring system.This kind of recognition of face crime scene investigation device, can control camera shooting
Valuable face is amplified in the electronic platform of head and the tracking of automatically controlled wide-angle lens, carries out optimum state candid photograph, is the backstage in a distant place
Recognition of face server or portable equipment provide the facial image of high value.This kind of recognition of face crime scene investigation device, can be in high-resolution
In the digital video image of rate or ultrahigh resolution, facial image is found out in real time, and sends backstage to and processing is identified.
In embodiment, recognition of face camera further includes being realized with picture of large image scale sensor using automatically controlled wide-angle lens
Electronic platform and automatically controlled wide-angle lens, can track on the image sensor, and face is captured in amplification.Controller, facial image are pre-
Processing and candid photograph module, memory and wireless communication module are set on circuit board, and circuit board is additionally provided with power supply and USB connects
Mouthful.Wireless communication module is communicated by telecommunications network with control platform.
This kind of recognition of face crime scene investigation device can be used for the people of matching identification to the image of real-time digital video after screening
The image data of face about only have original video data amount a ten thousandth, can effectively utilize existing telecommunication network or
Wireless network transmissions.And since recognition of face crime scene investigation device completes the work such as object identification and facial image screening, face is known
Other server, which can be sufficiently used for operational capability follow-up works, a servers such as face alignment identification, can handle multiple take the photograph
As the input picture of head, the fund input of expensive server is greatly saved.This kind of recognition of face crime scene investigation device, can solve people
The high resolution video image transmission problem and effective facial image Trapped problems encountered in face identification.
This kind of recognition of face crime scene investigation device, can find out and capture in huge vedio data has recognition of face
The facial image of value, to screen out it is a large amount of repeat, unwanted picture, provided for backstage recognition of face server important clear
Clear image, is greatly improved recognition speed and accuracy.
In embodiment, automatically controlled wide-angle lens accesses people with the collected real-time digital video of imaging sensor by controller
Face image pre-processes and captures module.After facial image pre-processes and capture module to real-time digital video pretreatment, filter out
It can be used for the image of the face of matching identification, or optimum state captured by controlling electronic platform and the tracking of automatically controlled wide-angle lens
Facial image, then facial image pretreatment and capture module by the image of the face that can be used for matching identification filtered out,
By image cache module cache and control platform is sent to by network interface.
Embodiment is integrated facial image pretreatment and candid photograph module and recognition of face camera.In such system
In, the collected real-time digital video access facial image pretreatment of camera and candid photograph module are pre- by real-time digital video
After processing, the image for the face that can be used for matching identification is filtered out, or control electronic platform and the tracking of automatically controlled wide-angle lens are grabbed
Best facial image is clapped, then these effective images, control platform is transferred to by telecommunications network, accesses face server pair
These images are identified.
In embodiment, automatically controlled wide-angle lens accesses 4K high-resolution image sensors, and imaging sensor is light image signal
It is converted into electric signal, by controller, image signal encoding is become per second 50 or 60 width digital videos, then accesses face
Image preprocessing and candid photograph module.Facial image pre-processes and captures the face in the every piece image of block search, if face
It is clear enough, there are enough pixels, just sends the face to the facial image cache module of rear class, be then sent to by network interface
Control platform.Control platform accesses recognition of face server and carries out face alignment.
In embodiment, such as Fig. 6, facial image pretreatment and candid photograph module carry out image preprocessing to real-time digital video,
Specifically, the foreground and background of original image is detached using differential technique, by the middle foreground image and background image of original image
It is detached, obtains foreground image, the boundary profile that convolution algorithm finds out object is carried out to foreground image and high-pass filtering template,
Independent object is isolated according to the continuity of profile and closure, the image of each object is normalized, then with
Skin detection carries out convolution algorithm, obtains subject image.
Facial image pre-processes and captures the image that module filters out the face that can be used for matching identification, specifically, right
After real-time digital video carries out image preprocessing, subject image is obtained, judges whether the subject image is front face, if
It is, then sees the clarity of facial image, if clarity is not less than setting value, filters out the face that can be used for matching identification
Image.
Facial image pre-processes and captures module and captures optimum state facial image, specifically, to real-time digital video
After carrying out image preprocessing, subject image is obtained, judges whether the subject image is front face, if so, seeing face figure again
The clarity of picture, if clarity is less than setting value, control electronic platform is directed at the face location, and controls automatically controlled wide-angle lens
Amplify and focus on the face, then according to facial image with positive eye opening skin detection convolution as a result, capture is best
The facial image of state.
Facial image is pre-processed and is captured in module, after carrying out image preprocessing to real-time digital video, obtains object figure
Picture judges whether the subject image is front face, is not front face if it is the subject image, then deletes the object figure
Picture.
In embodiment, high resolution video image filters big into after the facial image pretreatment for crossing front end and capturing module
The independent image data of amount, only screens and captures high-resolution facial image and be sent to distant place control platform, significantly reduces
The data volume of required transmission, while the facial image of high quality is provided, the burden of background server is alleviated, it can be meter
It calculates process resource and is used for recognition of face, improve recognition efficiency and speed.
In embodiment, facial image pretreatment and candid photograph module are integrated into recognition of face crime scene investigation device, are directly handled original
Digital video image rolls up foreground image and high-pass filtering template using the foreground and background of differential technique separate picture
The boundary profile of object is found out in product operation, independent object is isolated according to the continuity of profile and closure, to each object
Image be normalized, then with skin detection carry out convolution algorithm, judge whether the object is front face, such as
Fruit is the clarity for seeing facial image again, if clarity is inadequate, just controls electronic platform and is directed at the face location, and control electricity
Control wide-angle lens amplifies and focuses on the face, then according to the knot of facial image and positive eye opening skin detection convolution
Fruit captures optimum state facial image, is then passed to the further identifying processing of distant place background server.It can realize that processing is per second
The high resolution video image of 60 width 4K, in real time processing find any useful human face image information, Er Qieneng per piece image
Enough in the case of piece image up to 80 facial images, quickly controls electronic platform and automatically controlled wide-angle lens captures often respectively
A valuable people face image.
In embodiment, the high-definition digital image of 60 width 4K resolution ratio of real-time processing per second is realized using high speed ASIC, to figure
As carrying out calculus of differences wiping out background image, then carries out convolutional filtering and extract contour of object, multiple objects are isolated from image
The object separated is compared with face characteristic for body, excludes non-face object, thus people face from video image
It separates, up to 80 people faces can be discerned from each image.
After facial image is detected and separated from video, the angle of face and camera is judged according to face characteristic
Degree, picks out the face in face camera direction, controls the electronic platform tracking of camera, alignment, and adjust automatically controlled wide-angle lens
Head focuses, and amplifies the face, the facial image is captured in the case where face state is best.
A kind of mobile population video monitoring system based on recognition of face using above-mentioned recognition of face crime scene investigation device, such as Fig. 5,
Including front end surveillance device and control platform.
Front end surveillance device:Real-time digital video is collected by recognition of face crime scene investigation device, in real-time digital video
Original image is pre-processed, the image for the face that can be used for matching identification is filtered out, by the matching identification that can be used for of acquisition
The image of face sends control platform to by telecommunications network.
Control platform:The facial image transmitted by telecommunications network receiving front-end monitoring device, access face server is to this
A little facial images are identified.
Mobile population video monitoring system of this kind based on recognition of face, face identification system according to function and complicated journey
Degree is divided into front end surveillance device and control platform two parts:Front end surveillance device carries out high resolution digital video image preprocessing
With backstage recognition of face control platform so that front end can in the place that image generates, before being transmitted to Computer Vision,
Valuable facial image is filtered out, 99.9% or more hash is filtered, and controls electronic platform and automatically controlled wide-angle lens,
High-definition facial image is captured, only into excessively wired or wireless channel sends valuable image to the background service in a distant place
Device is handled and is identified into advancing one to facial image.
With reference to shown in figure 5 and Fig. 5, embodiment proposes handle for emerging recognition of face and the demand of artificial intelligence vision
The method that face snap and digital image video pretreatment are realized with special chip, and video pre-filtering be integrated into camera,
In crime scene investigation device and wearable device, after the high quality facial image shearing that then candid photograph is arrived, by wired or wireless network
It is sent to the system structure that facial image is compared identification in distant place control platform.In such a system, optical system and height
The digital video image that image in different resolution sensing acquisition arrives filters out valuable facial image, and control by processing in real time
Automatically controlled wide-angle lens and electronic platform track and capture the facial image of high quality, greatly reduce the data volume for needing to transmit,
The video processing load for sharing the backstage recognition of face server of distant place control platform, is greatly improved facial image recognition speed,
Accuracy and efficiency.
And than that described above, it is also necessary to which explanation is " one embodiment " spoken of in the present specification, " another reality
Apply example ", " embodiment " etc., refer to specific features, structure or the feature of embodiment description is combined to be included in the application general
In at least one embodiment of including property description.In the description multiple places occur statements of the same race be not centainly refer to it is same
Embodiment.Furthermore, it is understood that when describing a specific features, structure or feature in conjunction with any embodiment, what is advocated is
Realize that this feature, structure or feature are also fallen within the scope of the present invention in conjunction with other embodiment.
Although reference be made herein to invention has been described for multiple explanatory embodiments of the invention, however, it is to be understood that
Those skilled in the art can be designed that a lot of other modification and implementations, these modifications and implementations will be fallen in this Shen
It please be within disclosed scope and spirit.More specifically, disclose in the application, drawings and claims in the range of, can
With the building block and/or a variety of variations and modifications of layout progress to theme combination layout.In addition to building block and/or layout
Outside the variations and modifications of progress, to those skilled in the art, other purposes also will be apparent.
Claims (10)
1. a kind of recognition of face crime scene investigation device, it is characterised in that:Including recognition of face camera, memory and wireless communication module,
Recognition of face camera includes camera body, and camera body includes automatically controlled wide-angle lens, imaging sensor, controller and people
Face image pre-processes and captures module, and the optical signal of automatically controlled wide-angle lens is converted to digital signal by imaging sensor, and image passes
Sensor is by the digital data transmission of conversion to controller, and controller connects facial image pretreatment and captures module, and controller is also
It is separately connected memory and wireless communication module.
2. recognition of face crime scene investigation device as described in claim 1, which is characterized in that recognition of face camera further includes by automatically controlled wide
The electronic platform that angle mirror head is realized with imaging sensor, automatically controlled wide-angle lens project image onto on imaging sensor, face figure
It tracks as pre-processing and capturing module control electronic platform and amplifies face.
3. recognition of face crime scene investigation device as described in claim 1, which is characterized in that controller, facial image pretreatment and candid photograph
Module, memory and wireless communication module are set on circuit board, and circuit board is additionally provided with power supply and USB interface.
4. recognition of face crime scene investigation device as described in claim 1, which is characterized in that wireless communication module passes through telecommunications network and control
Platform communication.
5. recognition of face crime scene investigation device according to any one of claims 1-4, it is characterised in that:Automatically controlled wide-angle lens is passed with image
The collected real-time digital video of sensor accesses facial image pretreatment by high-definition digital video controller and captures module;People
After face image pre-processes and capture module to real-time digital video pretreatment, the figure for the face that can be used for matching identification is filtered out
Picture, or capture the facial image of optimum state, then facial image by controlling electronic platform and the tracking of automatically controlled wide-angle lens
Pretreatment and candid photograph module are cached the image of the face that can be used for matching identification filtered out by image cache module
And control platform is sent to by network interface.
6. recognition of face crime scene investigation device as claimed in claim 5, it is characterised in that:Facial image pre-processes and captures module to reality
When digital video carry out image preprocessing, specifically, using differential technique detach original image foreground and background, by original graph
The middle foreground image of picture is detached with background image, obtains foreground image, is rolled up to foreground image and high-pass filtering template
The boundary profile of object is found out in product operation, independent object is isolated according to the continuity of profile and closure, to each object
Image be normalized, then with skin detection carry out convolution algorithm, obtain subject image.
7. recognition of face crime scene investigation device as claimed in claim 5, it is characterised in that:Facial image pre-processes and captures module screening
Go out can be used for the image of the face of matching identification, specifically, after carrying out image preprocessing to real-time digital video, obtains object
Image judges whether the subject image is front face, if so, the clarity of facial image is seen again, if clarity is not low
In setting value, then the image for the face that can be used for matching identification is filtered out.
8. recognition of face crime scene investigation device as claimed in claim 5, it is characterised in that:Facial image pre-processes and captures module capture
Optimum state facial image obtains subject image specifically, after carrying out image preprocessing to real-time digital video, and judging should
Whether subject image is front face, if so, the clarity of facial image is seen again, if clarity is less than setting value, control
Electronic platform is directed at the face location, and controls automatically controlled wide-angle lens and amplify and focus on the face, then according to facial image
With positive eye opening skin detection convolution as a result, the facial image of capture optimum state.
9. recognition of face crime scene investigation device as claimed in claim 5, it is characterised in that:Facial image is pre-processed and is captured in module,
After carrying out image preprocessing to real-time digital video, subject image is obtained, judges whether the subject image is front face, if
Be the subject image it is not front face, then deletes the subject image.
10. recognition of face crime scene investigation device as claimed in claim 5, it is characterised in that:Control platform is received by telecommunications network and is transmitted
Facial image, access face server these facial images are identified.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810732512.5A CN108540707A (en) | 2018-07-05 | 2018-07-05 | Recognition of face crime scene investigation device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810732512.5A CN108540707A (en) | 2018-07-05 | 2018-07-05 | Recognition of face crime scene investigation device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108540707A true CN108540707A (en) | 2018-09-14 |
Family
ID=63487608
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810732512.5A Pending CN108540707A (en) | 2018-07-05 | 2018-07-05 | Recognition of face crime scene investigation device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108540707A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109670451A (en) * | 2018-12-20 | 2019-04-23 | 天津天地伟业信息系统集成有限公司 | Automatic face recognition tracking |
CN110175490A (en) * | 2018-09-21 | 2019-08-27 | 永康市巴九灵科技有限公司 | Game machine history account analysis system |
CN110784628A (en) * | 2019-08-14 | 2020-02-11 | 腾讯科技(深圳)有限公司 | Image data acquisition processing method and system, intelligent camera and server |
CN110969051A (en) * | 2018-09-29 | 2020-04-07 | 上海小蚁科技有限公司 | Face recognition method based on image sensor system and image sensor system |
CN112182008A (en) * | 2020-10-27 | 2021-01-05 | 青岛以萨数据技术有限公司 | System, method, terminal and medium for analyzing face picture data acquired by mobile terminal |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN201733373U (en) * | 2010-05-07 | 2011-02-02 | 北京英泰智软件技术发展有限公司 | High definition face capturing distinguishing camera |
CN205792930U (en) * | 2016-06-03 | 2016-12-07 | 广东万峯信息科技有限公司 | A kind of face snap machine and apply the monitoring system of this kind of face snap machine |
CN107278369A (en) * | 2016-12-26 | 2017-10-20 | 深圳前海达闼云端智能科技有限公司 | Method, device and the communication system of people finder |
CN107346426A (en) * | 2017-07-10 | 2017-11-14 | 深圳市海清视讯科技有限公司 | A kind of face information collection method based on video camera recognition of face |
CN208522852U (en) * | 2018-07-05 | 2019-02-19 | 郑永春 | Recognition of face crime scene investigation device |
-
2018
- 2018-07-05 CN CN201810732512.5A patent/CN108540707A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN201733373U (en) * | 2010-05-07 | 2011-02-02 | 北京英泰智软件技术发展有限公司 | High definition face capturing distinguishing camera |
CN205792930U (en) * | 2016-06-03 | 2016-12-07 | 广东万峯信息科技有限公司 | A kind of face snap machine and apply the monitoring system of this kind of face snap machine |
CN107278369A (en) * | 2016-12-26 | 2017-10-20 | 深圳前海达闼云端智能科技有限公司 | Method, device and the communication system of people finder |
CN107346426A (en) * | 2017-07-10 | 2017-11-14 | 深圳市海清视讯科技有限公司 | A kind of face information collection method based on video camera recognition of face |
CN208522852U (en) * | 2018-07-05 | 2019-02-19 | 郑永春 | Recognition of face crime scene investigation device |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110175490A (en) * | 2018-09-21 | 2019-08-27 | 永康市巴九灵科技有限公司 | Game machine history account analysis system |
CN110175490B (en) * | 2018-09-21 | 2021-04-16 | 泰州市津达电子科技有限公司 | Game machine historical account number analysis system |
CN110969051A (en) * | 2018-09-29 | 2020-04-07 | 上海小蚁科技有限公司 | Face recognition method based on image sensor system and image sensor system |
CN109670451A (en) * | 2018-12-20 | 2019-04-23 | 天津天地伟业信息系统集成有限公司 | Automatic face recognition tracking |
CN110784628A (en) * | 2019-08-14 | 2020-02-11 | 腾讯科技(深圳)有限公司 | Image data acquisition processing method and system, intelligent camera and server |
CN110784628B (en) * | 2019-08-14 | 2022-04-05 | 腾讯科技(深圳)有限公司 | Image data acquisition processing method and system, intelligent camera and server |
CN112182008A (en) * | 2020-10-27 | 2021-01-05 | 青岛以萨数据技术有限公司 | System, method, terminal and medium for analyzing face picture data acquired by mobile terminal |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108513110A (en) | Recognition of face monitoring camera | |
CN108540707A (en) | Recognition of face crime scene investigation device | |
WO2020078229A1 (en) | Target object identification method and apparatus, storage medium and electronic apparatus | |
US20130076943A1 (en) | Apparatus and method for image recognition of facial areas in photographic images from a digital camera | |
CN109711318B (en) | Multi-face detection and tracking method based on video stream | |
KR101798979B1 (en) | Smart Pedestrian Traffic Light System | |
US20230360254A1 (en) | Pose estimation method and related apparatus | |
CN112165573A (en) | Shooting processing method and device, equipment and storage medium | |
CN106778645A (en) | A kind of image processing method and device | |
CN109359577B (en) | System for detecting number of people under complex background based on machine learning | |
CN105022999A (en) | Man code company real-time acquisition system | |
CN108960146A (en) | The image pre-processing method of recognition of face monitor video | |
CN108921104A (en) | Mobile population video monitoring method based on recognition of face | |
WO2020249054A1 (en) | Living body detection method and system for human face by using two long-baseline cameras | |
KR101985711B1 (en) | Augmented reality CCTV system, and control method for the same | |
CN106897695A (en) | A kind of image recognizing and processing equipment, system and method | |
KR101621822B1 (en) | Device for security and recognizing number of license plate based on hybrid image transmission, and method for recognizing and providing number of license plate using the same | |
CN113673584A (en) | Image detection method and related device | |
CN111126411B (en) | Abnormal behavior identification method and device | |
CN108960147A (en) | Mobile population video monitoring system based on recognition of face | |
Kaur et al. | Cloud Based Surveillance using ESP32 CAM | |
CN208522852U (en) | Recognition of face crime scene investigation device | |
KR101337554B1 (en) | Apparatus for trace of wanted criminal and missing person using image recognition and method thereof | |
CN208691422U (en) | Recognition of face monitoring camera | |
CN110324589A (en) | A kind of monitoring system and method for tourist attraction |
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 |