CN106372606A - Target object information generation method and unit identification method and unit and system - Google Patents

Target object information generation method and unit identification method and unit and system Download PDF

Info

Publication number
CN106372606A
CN106372606A CN201610798111.0A CN201610798111A CN106372606A CN 106372606 A CN106372606 A CN 106372606A CN 201610798111 A CN201610798111 A CN 201610798111A CN 106372606 A CN106372606 A CN 106372606A
Authority
CN
China
Prior art keywords
destination object
data
described destination
information
characteristic information
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
Application number
CN201610798111.0A
Other languages
Chinese (zh)
Inventor
程凯
印奇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Megvii Technology Co Ltd
Beijing Aperture Science and Technology Ltd
Original Assignee
Beijing Megvii Technology Co Ltd
Beijing Aperture Science and Technology Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Beijing Megvii Technology Co Ltd, Beijing Aperture Science and Technology Ltd filed Critical Beijing Megvii Technology Co Ltd
Priority to CN201610798111.0A priority Critical patent/CN106372606A/en
Publication of CN106372606A publication Critical patent/CN106372606A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
    • G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
    • G06F16/5866—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using information manually generated, e.g. tags, keywords, comments, manually generated location and time information
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
    • G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
    • G06F16/583—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/22—Matching criteria, e.g. proximity measures
    • 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
    • G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/94—Hardware or software architectures specially adapted for image or video understanding
    • G06V10/95—Hardware or software architectures specially adapted for image or video understanding structured as a network, e.g. client-server architectures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/40—Scenes; Scene-specific elements in video content
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/40—Scenes; Scene-specific elements in video content
    • G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
    • 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
    • G06V20/54—Surveillance or monitoring of activities, e.g. for recognising suspicious objects of traffic, e.g. cars on the road, trains or boats
    • 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
    • G06V40/161—Detection; Localisation; Normalisation
    • 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
    • G06V40/168—Feature extraction; Face representation
    • G06V40/171—Local features and components; Facial parts ; Occluding parts, e.g. glasses; Geometrical relationships
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/60—Type of objects
    • G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
    • G06V20/625—License plates

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Data Mining & Analysis (AREA)
  • Health & Medical Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Library & Information Science (AREA)
  • General Health & Medical Sciences (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Biology (AREA)
  • Databases & Information Systems (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • Human Computer Interaction (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Software Systems (AREA)
  • Image Analysis (AREA)

Abstract

The invention discloses a target object information generation method and unit, an identification method and unit and system. According to the target object information generation method, datamation and formatting of the characteristic information of a target object are accomplished at a camera end, and the characteristic information after datamation is transmitted to a server through a special data format so as to enable the server to identify the target object according to the characteristic information after datamation. Through the target object information generation method and unit, and the identification method and unit and system, network bandwidth and resources occupied by the target object during transmission can be reduced, transmission efficiency and timeliness are improved, operation complexity of a target object identification process is reduced, identification efficiency is improved, and business identification demands of large-scale target objects can be satisfied.

Description

Target object information generation method and unit, recognition methodss and unit and system
Technical field
The present invention relates to image identification technical field, more particularly, to one kind can reduce transmitted data on network amount, improve mesh A kind of recongnition of objects method of mark Object identifying efficiency, target object information generation method, target object information generate single Unit, recongnition of objects unit and system.
Background technology
With expanding economy, advancing by leaps and bounds of the quickening of city-building speed, and the Internet, lead to population in city close Collection, recurrent population increases, and has caused the traffic in urban construction, social security, key area is taken precautions against, the network crime becomes increasingly conspicuous Deng city management problem.In the boundless and indistinct sea of faces, how to be accurately positioned suspected target, always public security is intelligent, informatization The most important thing.
Artificial intelligence's introducing monitoring field is also become trend, the particularly monitoring system of recognition of face to security protection stability maintenance, Arrest runaway convict, there is the great market demand in the field such as identification, estimate to I haven't seen you for ages and form the industry of hundred billion RMB ranks Chain.How to design the rational face identification system of a set of high-quality to this industry realize early undoubtedly have huge motive force.
Current destination object (for example, face) identifying system mainly comprises two parts: front-end camera and be used for target The back-end server of Object identifying.What front-end camera was transferred to back-end server is video flowing, so needing to take high Bandwidth, can cause bandwidth pressure big or the problems such as poor transmission, and the core work amount of recongnition of objects (is scratched figure and joined Examine image to compare) to be all back-end server to complete, and this leads to the operand of back-end server greatly, does not meet what distribution calculated Framework thinking, under efficiency is relatively low.
In general, there are two defects in existing recongnition of objects system: 1) network bandwidth requirement is higher, for complexity Network environment can have such problems as time delay, the real-time of identification cannot ensure；2) business of server is complex, tailored version Still not strong it is impossible to meet the requirement of extensive recongnition of objects business, or even easily occur delaying machine the problems such as.
Accordingly, it is desirable to provide a kind of recognition methodss of destination object, unit and system, to solve at least in part to carry above The problem arriving.
Content of the invention
Introduce a series of concept of reduced forms in Summary, this will specific embodiment partly in enter one Step describes in detail.The Summary of the present invention is not meant to the pass attempting to limit technical scheme required for protection Key feature and essential features, more do not mean that the protection domain attempting to determine technical scheme required for protection.
In order to solve the above problems at least in part, the invention provides a kind of recongnition of objects method, comprising: receive The format data of described destination object, the format data of described destination object is to comprise described target by what video camera generated The digitization characteristic information of object；The format data of object will be compared in the format data of described destination object and data base Carry out data retrieval comparison；Determined in described data base with the presence or absence of consistent with described destination object based on retrieval comparison result Compare object, if exist in described data base consistent with described destination object compare object, transfer and described target pair As the consistent identity information comparing corresponding to object.
Exemplarily, the format data of described destination object is obtained by described video camera execution following steps: collection bag Include the image of destination object；Destination object is extracted from described image；The feature of destination object is extracted from described destination object Information；By the characteristic information dataization of described destination object and formatting to form the format data of described destination object.
Exemplarily, the format data of described destination object is the digitization characteristic information of described destination object to set The array data of data form arrangement form.
Exemplarily, the digitization characteristic information of described destination object include data head, video camera information stamp, timestamp, Address stamp, key point information, data tail.
Exemplarily, each number in carrying out the format data to described destination object when described data retrieval compares Carry out parallel search comparison according to changing characteristic information.
Exemplarily, carrying out providing similarity evaluation for each described comparison object when described data retrieval compares Value.
Exemplarily, each in carrying out the format data based on described destination object when described data retrieval compares Compare in digitization characteristic information and data base in the format data of format data and described destination object of object The similarity of each digitization characteristic information corresponding digitization characteristic information is directed to each described comparison object and provides similarity Evaluation of estimate.
Exemplarily, based on each described compare object similarity evaluation value determine in described data base whether there is with The consistent comparison object of described destination object, wherein, sets when the similarity evaluation value that at least there is a comparison object is more than During threshold value, what in the described data base of judgement, presence was consistent with described destination object compares object, and selects maximum similarity evaluation Value is corresponding, and to compare object be consistent with described destination object to compare object；Conversely, then judging do not exist in described data base Consistent with described destination object compares object.
According to the recongnition of objects method of the present invention, by receiving the format data of described destination object, and it is based on The formatted data comparing object in the format data of this destination object and data base carries out data retrieval and compares determining No have consistent comparison object, and directly carries out compared with image retrieval compares, due to only needing transmission formatting data, thus fall The low data volume of network transmission, and compare, due to line retrieval is entered using format data, the complexity reducing identification business And treating capacity, improve recognition efficiency it is ensured that the real-time of recongnition of objects.
According to another aspect of the present invention, additionally provide a kind of target object information generation method, comprising: pass from image Detect in the image of sensor collection and extract destination object；Extract the characteristic information of described destination object；To described destination object Characteristic information carry out digitization and formatting, comprise the format data of described destination object characteristic information to be formed；Wherein, Abovementioned steps are all executed by video camera.
Exemplarily, detect and extract destination object the described image from imageing sensor collection and include: from described figure As detecting the coordinate data of described destination object position in the image of sensor acquisition；Coordinate according to described destination object Data, takes, from the image of described image sensor collection, the image only including described destination object.
Exemplarily, the format data of described destination object is the digitization characteristic information of described destination object to set The array data of data form arrangement form.
Exemplarily, the digitization characteristic information of described destination object include data head, video camera information stamp, timestamp, Address stamp, key point information, data tail.
According to the target object information generation method of the present invention, by the characteristic information of destination object is formatted place Reason, to form the format data of destination object, so when spreading out of the format data of this destination object by network, just The data volume of network transmission can be substantially reduced, save the network bandwidth and resource, improve the real-time of destination object characteristic information transmission Property.
According to another aspect of the present invention, additionally provide a kind of target object information signal generating unit it is characterised in that institute State target object information signal generating unit to be integrated in video camera, and described target object information signal generating unit includes: target pair As Detection and Extraction module, described destination object Detection and Extraction module is used for detecting from the image of imageing sensor collection and extracting Destination object；Characteristic information extracting module, described characteristic information extracting module extracts destination object from described destination object Characteristic information；Format data generation module, described format data generation module is used for the feature letter of described destination object Breath digitization and formatting, to form the format data comprising described destination object characteristic information.
Exemplarily, described destination object Detection and Extraction module includes: coordinate measurement submodule, for passing from described image The coordinate data of described destination object position is detected in the image of sensor collection；Image takes submodule, for according to institute State the coordinate data of destination object, take, from the image of described image sensor collection, the figure only including described destination object Picture.
Exemplarily, the format data of described destination object is the digitization characteristic information of described destination object to set The array data of data form arrangement form.
Exemplarily, the digitization characteristic information of described destination object include data head, video camera information stamp, timestamp, Address stamp, key point information, data tail.
According to the target object information signal generating unit of the present invention, by the characteristic information of destination object is formatted place Reason, to form the format data of destination object, so when spreading out of the format data of this destination object by network, just The data volume of network transmission can be substantially reduced, save the network bandwidth and resource, improve the real-time of destination object characteristic information transmission Property.
According to another aspect of the present invention, additionally provide a kind of recongnition of objects unit, comprising: information receives mould Block, described information receiver module is used for receiving the format data of destination object, the formatting of described destination object by network Data is the digitization characteristic information comprising described destination object being generated by video camera；Retrieval comparing module, described retrieval ratio Module is used for the format data comparing object in data base, the format data of described destination object is carried out data Retrieval compares, and compares object to determine in described data base with the presence or absence of consistent with described destination object；Processing module, described Processing module described retrieval comparing module determine exist in described data base consistent with described destination object when comparing object, Transfer the identity information that compare object corresponding to consistent with described destination object.
Exemplarily, the format data of described destination object is the digitization characteristic information of described destination object to set The array data of data form arrangement form.
Exemplarily, the digitization characteristic information of described destination object include data head, video camera information stamp, timestamp, Address stamp, key point information, data tail.
Exemplarily, digitization feature letter in the format data to described destination object for the described retrieval comparing module Breath carries out parallel search comparison.
Exemplarily, described retrieval comparing module is carrying out being directed to each described comparison object when described data retrieval compares Provide similarity evaluation value.
Exemplarily, described retrieval comparing module is carrying out the lattice based on described destination object when described data retrieval compares Format data the and described destination object of object is compared in each digitization characteristic information in formula data and data base Format data in each digitization characteristic information corresponding digitization characteristic information similarity be directed to each described ratio Object is given with similarity evaluation value.
Exemplarily, described retrieval comparing module determines described number based on each described Similarity value evaluation comparing object Compare object according in storehouse with the presence or absence of consistent with described destination object, wherein, described retrieval comparing module at least has one When individual similarity evaluation value is more than the comparison object of given threshold, judge that in described data base, presence is consistent with described destination object Comparison object, and select maximum similarity evaluation of estimate corresponding compare object be consistent with described destination object compare right As；When not depositing the comparison object that similarity evaluation is more than given threshold, then judge do not exist in described data base and described mesh The consistent comparison object of mark object.
According to the recongnition of objects unit of the present invention, by receiving the format data of described destination object, and it is based on The formatted data comparing object in the format data of this destination object and data base carries out data retrieval and compares determining No have consistent comparison object, and directly carries out compared with image retrieval compares, due to only needing transmission formatting data, thus fall The low data volume of network transmission, and compare, due to line retrieval is entered using format data, the complexity reducing identification business And treating capacity, improve recognition efficiency it is ensured that the real-time of recongnition of objects.
According to another aspect of the present invention, additionally provide a kind of recongnition of objects system, comprising: image sensing Device, described image sensor is used for gathering image；Target object information signal generating unit, described target object information signal generating unit is used Acquisition destination object in the image gathering from described image sensor, and extract the characteristic information of institute's destination object, and simultaneously Digitization and formatting are carried out to described destination object characteristic information, to form the format data of described destination object；Target Object identification unit, described recongnition of objects unit is used for receiving the format data of described destination object, and by described mesh The format data comparing object in format data and the data base of mark object carries out data retrieval and compares, to determine described number Compare object according in storehouse with the presence or absence of consistent with described destination object, and carry out respective handling, wherein, described destination object is known Other unit exists consistent with described destination object when comparing object in determining described data base, transfers and described destination object The consistent identity information comparing corresponding to object；There is not the ratio consistent with described destination object in determining described data base During to object, instruction proceeds the identification of next destination object；Communication unit, described communication unit is configured to described target The format data of the described destination object that object information signal generating unit generates sends to described recongnition of objects unit.
Exemplarily, described image sensor, target object information signal generating unit and communication unit configure in distal view picture In harvester, described recongnition of objects unit configures in server end.
Exemplarily, the quantity of described far-end image collecting device and described server end is than for n:1, wherein n be more than Natural number equal to 1.
Exemplarily, the format data of described destination object is the digitization characteristic information of described destination object to set The array data of data form arrangement form.
Exemplarily, the digitization characteristic information of described destination object include data head, video camera information stamp, timestamp, Address stamp, key point information, data tail.
Generated single according to the recongnition of objects method of the present invention, target object information generation method, target object information Unit, recongnition of objects unit and system, on the one hand, complete to gather the image of destination object, extract target pair by video camera The characteristic information of elephant, and form the format data of destination object based on this feature information, and by network by this target pair The format data of elephant is sent to server (or high in the clouds), so due to only need send format data, therefore and between send out Send video or picture to compare the data volume greatly reducing network transmission, improve the real-time of network transmission, and saved net Network bandwidth and resource；On the other hand pass through in the format data based on this destination object for the server (or high in the clouds) and data base The formatted data of comparison object carry out data retrieval and compare determining whether there is consistent comparison object, so with directly enter Row image retrieval compares and compares, and due to the retrieval retrieved similar to conventional characters data of format data, operand drops significantly Low, thus reduce complexity and the treating capacity of identification business, improve recognition efficiency it is ensured that recongnition of objects real-time Property, it is suitable to meet the requirement of extensive recongnition of objects business.
Brief description
The drawings below of the embodiment of the present invention is used for understanding the present invention in this as the part of the present invention.Shown in accompanying drawing Embodiments of the invention and its description, for explaining the principle of the present invention.In the accompanying drawings,
Fig. 1 is to know for realizing target object information generation method according to embodiments of the present invention and device, destination object The schematic block diagram of the exemplary electronic device of other method and apparatus；
Fig. 2 is the indicative flowchart of the recongnition of objects method according to the embodiment of the present invention；
Fig. 3 is the indicative flowchart of the target object information generation method according to the embodiment of the present invention；
Fig. 4 is to compare a schematic flow according to data retrieval in the recongnition of objects method of the embodiment of the present invention Figure；
Fig. 5 is to compare another schematic flow according to data retrieval in the recongnition of objects method of the embodiment of the present invention Figure；
Fig. 6 is the schematic block diagram of the target object information signal generating unit according to the embodiment of the present invention；
Fig. 7 is the schematic block diagram of the recongnition of objects unit according to the embodiment of the present invention；
Fig. 8 is the schematic block diagram of the recongnition of objects system according to the embodiment of the present invention；
Fig. 9 illustrates a kind of Face datection and stingy diagrammaticalIy schematic diagram at present；
Figure 10 illustrates a kind of face alignment diagrammatic view in principle at present.
Specific embodiment
In the following description, a large amount of concrete details are given to provide more thorough understanding of the invention.So And, it will be apparent to one skilled in the art that the embodiment of the present invention can one or more of these details And be carried out.In other examples, in order to avoid obscuring with the embodiment of the present invention, for more well known in the art Technical characteristic is not described.
It should be appreciated that the present invention can be implemented in different forms, and should not be construed as being limited to propose here Embodiment.On the contrary, it is open thoroughly and complete to provide these embodiments will make, and will fully convey the scope of the invention to Those skilled in the art.In the accompanying drawings, in order to clear, the size of part, element etc. and relative size may be exaggerated.Ad initio Represent identical element to whole same reference numerals.
So that the object, technical solutions and advantages of the present invention become apparent from, describe root below with reference to accompanying drawings in detail Example embodiment according to the present invention.Obviously, described embodiment is only a part of embodiment of the present invention, rather than this Bright whole embodiments are not it should be appreciated that the present invention is limited by example embodiment described herein.Described in the present invention The embodiment of the present invention, the obtained all other embodiment in the case of not paying creative work of those skilled in the art All should fall under the scope of the present invention.
Current face recognition process mainly includes following two processes；First, as shown in figure 9, detecting in video flowing Each two field picture in facial image and provide the coordinate of image, then by coordinate, face head portrait is plucked out；Then, as Figure 10 institute Show, stingy figure carried out aspect ratio pair with the face picture of face database, when reaching certain threshold value, that is, is considered same person, Thus recalling the identity information of this person, and then finally realize face identification functions.In this recognition methodss, directly adopt face Image enters line retrieval and compares, and identification operand is very big, and identification real-time is relatively low, and is difficult to meet extensive recognition of face business Require.
The present invention, in order to improve the recognition efficiency of destination object, reduces identification business complexity and operand, to target pair The characteristic information of elephant has carried out digitization and formatting is processed, and the immediate data retrieval when comparing compares, thus significantly drops Low operand.And it is preferable that in an embodiment of the present invention, using Distributed Design, that is, in front end (for example, video camera End) digitization of destination object characteristic information and formatting are processed, obtain the formatting number comprising destination object characteristic information According to, carry out data retrieval comparison in rear end (for example, server or high in the clouds), thus both reduced network transmission data volume and Bandwidth requirement, improves the recognition efficiency that object recognition equipment is marked in rear end (for example, server or high in the clouds) again, to meet big rule The requirement of mould recognition of face business.
In the present invention, so-called destination object refers to target that is various to be identified or being identified, the machine such as such as automobile The license plate image of motor-car, facial image of various crowd etc..In other words, the recongnition of objects side in the present invention Method, target object information generation method can be applied particularly to Car license recognition, field of face identification, or are based on Car license recognition, people Field of video monitoring of face identification etc..It is, of course, understood that described destination object is not limited to car plate and face, and can To be various recognizable object.And destination object herein is implication described herein, will not be described in great detail later.
Below by taking recognition of face as a example to the recongnition of objects method of the present embodiment, target object information generation side Method, target object information signal generating unit, recongnition of objects unit and system are described.It should be appreciated that these of the present invention Method, device and system all can apply to the identification of other destination objects, acquisition of information etc., and recognition of face is only one and shows Meaning property application example.
First, to describe with reference to Fig. 1 for realizing target object information generation method according to embodiments of the present invention and dress Put, the exemplary electronic device 100 of recongnition of objects method and apparatus.
As shown in figure 1, electronic equipment 100 includes one or more processors 102, one or more storage device 104, defeated Enter/output device 106, communication interface 108 and one or more imageing sensor 110, these assemblies pass through bus system 112 And/or bindiny mechanism's (not shown) interconnection of other forms.It should be noted that the assembly of electronic equipment 100 shown in Fig. 1 and structure Simply illustrative, and not restrictive, as needed, described electronic equipment can also have other assemblies and structure, also may be used Not include aforesaid members, for example, can include imageing sensor 110 it is also possible to not include imageing sensor 110.
Described processor 102 typically represent any types or form can processing data or explain and execute instruction place Reason unit.In general, processor can be CPU (cpu) or have data-handling capacity and/or instruction hold The processing unit of the other forms of row ability, and the other assemblies in described electronic equipment 100 can be controlled to execute expectation Function.In a particular embodiment, processor 102 can receive the instruction from software application or module.These instructions are permissible Lead to the function of one or more example embodiment that processor 102 completes to be described herein and/or illustrate.
Described storage device 104 can include one or more computer programs, and described computer program can To include various forms of computer-readable recording mediums, such as volatile memory and/or nonvolatile memory.Described easy The property lost memorizer for example can include random access memory (ram) and/or cache memory (cache) etc..Described non- Volatile memory for example can include read only memory (rom), hard disk, flash memory etc..In described computer-readable recording medium On can store one or more computer program instructions, processor 102 can run described program instruction, to realize hereafter institute The client functionality (realized by processor) in the embodiment of the present invention stated and/or other desired function.In described meter Various application programs and various data can also be stored in calculation machine readable storage medium storing program for executing, such as described application program using and/or Various data producing etc..
Described input/output device 106 can be the device for input instruction with to the various information of outside output for the user, For example input equipment can include one or more of keyboard, mouse, mike and touch screen etc..Output device can include One or more of display, speaker etc..
Communication interface 108 widely represent any types or form can promote exemplary electronic device 100 and one or The communication equipment of the communication between multiple optional equipments or adapter.For example, communication interface 108 can promote electronic equipment 100 Communication with front end or accessory electronic device and back-end server or high in the clouds.The example of communication interface 108 includes but is not limited to Wired network interface (such as NIC), radio network interface (such as wireless network interface card), modem and appoint What his suitable interface.In one embodiment, communication interface 108 by with the direct-connected offer of the network of such as the Internet to remote Journey server/remote front-end equipment direct-connected.In a particular embodiment, communication interface 108 by with dedicated network, such as video The direct-connected offer of the networks such as monitoring network, Skynet system network is direct-connected to remote server/remote front-end equipment.Communication interface 108 can also provide this connection by any other suitable connection indirectly.
Described image sensor 110 can shoot the desired image of user (such as photo, video etc.), and will be captured Image be stored in described storage device 104 for other assemblies use.
Exemplarily, for realizing target object information generation method according to embodiments of the present invention and device, target pair As the exemplary electronic device of recognition methodss and device may be implemented as the figure of such as smart mobile phone, panel computer, gate control system As collection terminal, the image acquisition end of preventing road monitoring system, the image acquisition end of safety-protection system and various monitoring, safety-security area etc. Rear end control process end or server end or high in the clouds etc..
Fig. 2 is the indicative flowchart of the recongnition of objects method according to the embodiment of the present invention.Below in conjunction with Fig. 2 pair Recongnition of objects method according to embodiments of the present invention is described.
As shown in Fig. 2 recongnition of objects method includes disclosed in the embodiment of the present invention:
First, the format data of described destination object, the formatting number of described destination object in step s201, are received According to the digitization characteristic information comprising described destination object for being generated by video camera.
The format data of described destination object is that the characteristic information based on destination object is formed with fixed data form Data, it comprises the digitization characteristic information of described destination object.Fixed data form be, for example, array data, set and Other data types or self-defining data type.By by the characteristic information data of destination object and formatting, permissible Reduce the network bandwidth shared by transmission information and flow, improve network transmission efficiency and real-time.
Exemplarily, because the operand of destination object analysis identification is relatively large, therefore, recongnition of objects method can With in server end or high in the clouds execution, thus improving recognition speed and ensureing the real-time identifying.
In embodiments of the present invention, the format data of destination object is generated by video camera, and gives clothes by camera transmissions Business device, directly transmits picture format in hinge structure, reduces bandwidth occupancy rate, accelerates transmission speed, and then ensures identification Real-time.Exemplarily, the format data of described destination object is obtained by described video camera execution following steps: collection includes The image of destination object；Destination object is extracted from described image；The feature letter of destination object is extracted from described destination object Breath；By the characteristic information dataization of described destination object and formatting to form the format data of described destination object.
In step s202, carry out data retrieval and compare with the object that compares in data base.
The formatting number of object is compared in the format data of destination object that will receive in step s201 and data base According to carry out data retrieval comparison.Compare due to directly line retrieval being entered using format data when retrieval compares, scheme with adopting Piece retrieval comparison is compared, and operand substantially reduces, and recognition efficiency improves, and is suitable to meet the extensive requirement identifying business.
In step s203, determined in described data base based on retrieval comparison result and whether there is and described destination object one The comparison object causing.
If there is the compare object consistent with described destination object in described data base, entering in step s204, adjusting Take the identity information that compare object corresponding to consistent with described destination object.
Identity information such as name, household register, working cell, house address and the letter such as semantic object extraction time and address Breath is sent to upper system, and such as public security system, warning system, road traffic centring system etc. carry out subsequent treatment.
If there is not the compare object consistent with described destination object in described data base, enter in step s205.
In step s205, judge recognition failures, proceed the identification of next destination object.
In some instances, when carrying out data retrieval comparison in step s202, can be for each ratio in data base Object is given with similarity evaluation value.
In other examples, step s202 carries out the lattice based on described destination object when described data retrieval compares Format data the and described destination object of object is compared in each digitization characteristic information in formula data and data base Format data in each digitization characteristic information corresponding digitization characteristic information similarity be directed to each described ratio Object is given with similarity evaluation value.Exemplarily, by each feature of destination object, such as on face overall information, face Nevuss, left ear information, auris dextra information etc. compare the corresponding informance of object and enter line retrieval and compare with data base respectively, draw and appoint One comparison object each feature similarity it is assumed that destination object with any one compare object face overall information phase Seemingly spending the similarity for the nevuss on a, face is that b, left ear information and auris dextra information similarity are respectively c and d, based on a, b, c and d Similarity evaluation value is provided to this comparison object, for example, it is possible to set a weighted value for each similarity it is assumed that distinguishing in advance For a, b, c and d, the similarity evaluation value to this comparison object can be a × a+b × b+c × c+d × d.
Further, determined in described data base based on this similarity evaluation value in step s203 and whether there is and described mesh The consistent comparison object of mark object.For example when the respective value of format data is close to (for example, aforementioned a, b, c and d respective value is all relatively When greatly), then provide higher similarity evaluation value, when the respective value difference of format data is larger, then provide relatively low phase Evaluate like degree.Which kind of which kind of as scope to be larger for difference close to scope in, then can be set according to concrete identification design Fixed, here is not particularly limited.
Further, when the similarity evaluation value that at least there is a comparison object is more than given threshold, judge described What in data base, presence was consistent with described destination object compares object, thus transferring corresponding identity information；Conversely, then judging institute State and in data base, there is not the formatting that compare object, can reacquire described destination object consistent with described destination object Data, and carry out abovementioned steps 202 to step 204, or the comparison that next destination object can be proceeded.
Exemplarily, determined in described data base based on each described similarity evaluation comparing object and whether there is and institute State the consistent comparison object of destination object, such as when the similarity evaluation value that at least there is a comparison object is more than given threshold When, what in the described data base of judgement, presence was consistent with described destination object compares object, and selects maximum similarity evaluation of estimate pair The object that compares answered is consistent with described destination object to compare object；Conversely, then judging do not exist in described data base and institute State the consistent comparison object of destination object.
Exemplarily, the digitization characteristic information of destination object includes data head, video camera information stamp, timestamp, address Stamp, key point information, data tail.As shown in Table 1 below:
Table 1
As shown in table 1, be the digitization characteristic information of the destination object of exemplary embodiment of the present schematic diagram, its In, the heading of data head such as tcp message, the port numbers of source port and destination interface, serial number seq, response can be included Number ack, attribute field etc., still taking tcp message as a example, by digitization characteristic information for example video camera information stamp, timestamp, Location stamp, key point information etc. are encapsulated in tcp message, exemplarily, are encapsulated in tcp message with setting data form, such as table 1 Shown in, in the data flow of tcp message, digitization characteristic information put in order as video camera information stamp, information stamp and time Stab, be key point information again, wherein, the arrangement of key point information can be face overall information such as shape of face, face cloth first The information such as office, are then nevuss on such as face, freckle, at least one characteristic information such as scar, then for left ear information, the right side again Ear information, left eyebrow information, right eyebrow information, left eye information, right eye information, nose information and face information etc. at least one.By number It is set by data form arrangement according to the characteristic information changed to contribute to accelerating speed when retrieval compares, thus improving identification effect Rate, for example, the server for recongnition of objects can quickly identify, by video camera information stamp, the letter knowing this video camera Breath, can quickly position shape of face, face layout of current target object etc. by face overall information.Key point information can be wrapped Include at least one in the various information being previously mentioned it is to be understood that when including aforementioned all information, its identification is accurately Rate is higher it is adaptable to some high accuracy identify scenes.
Based on the form changing characteristic information as shown by the data in table 1, the present invention also provides another kind of exemplary embodiment, first will Destination object is compared with arbitrary face overall information (such as shape of face and face layout) comparing object, only when it is similar When degree is more than predetermined threshold value, for example, during more than 80%, just carry out the comparison of other characteristic informations further, otherwise will be with next Individual comparison object is compared, and so constantly drawdown ratio to scope, can improve recognition speed.First carry out with face overall information Comparing and be only an example, can first other any feature information being compared, here of the present invention does not limit, as long as meeting When the similarity comparing with arbitrary any feature information comparing object is more than predetermined threshold value constantly, then carry out comparing object with this The comparison of next characteristic information design philosophy.
Exemplarily, the digitization characteristic information in the embodiment of the present invention can be understood as by characteristic information such as human eye, The characteristics such as human ear, vectorization, for example, represent the type of characteristic information, the coordinate of characteristic information by binary data Deng.Its little, occupied bandwidth of taking up room of characteristic information after digitization is little, and transmission speed is fast, improves the real-time of identification.
Exemplarily, the recongnition of objects method according to the present embodiment can have setting of memorizer and processor Realize in standby, device or system.Recongnition of objects method according to embodiments of the present invention can be deployed in facial image and know At other identification end, control process end of gate control system of the master control end of such as preventing road monitoring system, Companies House etc. etc..
According to the recongnition of objects method of the present embodiment, by receiving the format data of described destination object, and base The formatted data comparing object in format data and the data base of this destination object carries out data retrieval and compares to determine With the presence or absence of consistent comparison object, and directly carry out compared with image retrieval compares, due to only needing transmission formatting data, thus Reduce the data volume of network transmission, and compare, due to line retrieval is entered using format data, the complexity reducing identification business Degree and treating capacity, improve recognition efficiency it is ensured that the real-time of recongnition of objects.
Fig. 3 is the indicative flowchart of the target object information generation method according to the embodiment of the present invention.With reference to Fig. 3 Target object information generation method according to embodiments of the present invention is described, wherein, in target object information generation method All steps executed by video camera or image collecting device, video camera or image collecting device include image sensing Device, imageing sensor is used for gathering image.
As shown in figure 3, target object information generation method includes disclosed in the embodiment of the present invention:
First, in step s301, detect from the image of imageing sensor collection and extract described destination object.
The image of described image sensor collection can also be able to be video for picture.Image when imageing sensor collection During for video, then each frame picture being directed to this video of composition carries out detection and the extraction of destination object.In this step, permissible Determine whether comprise destination object in imageing sensor acquired image, and in imageing sensor acquired image In comprise destination object in the case of orient the region of destination object in imageing sensor acquired image.
Taking recognition of face as a example, then can determine that whether comprise face in imageing sensor acquired image, and And orient in imageing sensor acquired image in the case of comprising face in imageing sensor acquired image Human face region.Exemplary property, for example can be using the good human-face detector of training in advance come the figure being gathered in imageing sensor Locating human face region in picture.For example, it is possible to advance with the human face detection and tracing such as Ha Er (haar) algorithm, adaboost algorithm Algorithm trains human-face detector on the basis of a large amount of pictures, for the single-frame imagess of input, the good face of this training in advance Detector can quickly locate out human face region.Additionally, for the multiple image of imageing sensor continuous acquisition, in first frame figure As in orient human face region after, can position based on human face region in the previous frame image of current frame image come in real time Follow the trail of the position of human face region in current frame image.
After detecting facial image and provide corresponding coordinate in the image of imageing sensor collection, then pressing coordinate afterwards will Facial image plucks out (being commonly called as scratching figure), to obtain facial image.
Exemplarily, described destination object can be detected from the image of described image sensor collection in step s301 The coordinate data of position；According to the coordinate data of described destination object, scratch from the image of described image sensor collection Take the image only including described destination object.As an example, after determining the coordinate data of destination object position, can To be marked at these coordinates, for example, it is labeled as 1, and the position mark outside these coordinates is 0, so that it is determined that destination object Region, profile and pixel etc., then can will be labeled as 1 with existing figure cutting techniques (for example, graph cut algorithm) Region takes out, thus only being comprised the image of destination object (for example, face).
In step s302, extract the characteristic information of described destination object.That is, the destination object extracting from step s301 In extract the characteristic information of destination object further.
So-called characteristic information, refer to can be used for identification object or is distinguished destination object and other objects Information.
, so-called characteristic information exemplarily can include such as face overall information taking facial image as a example, such as shape of face, Face layout etc., can also include local feature information, such as nevuss, freckle, scar etc., can also include face characteristic information, For example left ear information (size, shape etc.), auris dextra information (size, shape etc.), left eyebrow information (such as shape, position, length and width Deng), right eyebrow information (such as shape, position, length and width etc.), and left eye information (such as shape, position, length and width etc.), right eye Information (such as shape, position, length and width etc.) etc..
Exemplarily, in one embodiment, sobel operator extraction characteristics algorithm or gabor characteristics algorithm etc. can be passed through The characteristic information of facial image is extracted in the facial image that feature extraction algorithm extracts from step 301.In another embodiment, Can be using the good crucial spot locator of training in advance come locating human face's key point in human face region (such as eyes, canthus, eye Eyeball center, eyebrow, nose, nose, face, the corners of the mouth and face mask etc.).For example, it is possible to advance with cascade homing method exist Crucial spot locator is trained in a large number on the basis of the face picture of artificial mark.Alternatively, it would however also be possible to employ traditional face Crucial independent positioning method, it is based on parametric shape model, according to the appearance features near key point, learns a parameter mould Type, iteratively optimizes the position of key point, finally obtains crucial point coordinates when using, after obtaining crucial point coordinates, permissible One step is carried out based on this coordinate and obtains key point (such as eyes, canthus, eye center, eyebrow, nose, nose, face, the corners of the mouth With face mask etc.) other information, such as size, shape etc. information.
Should be appreciated that the present invention is not limited by the method for detecting human face specifically adopting and characteristics information extraction method, either Existing method for detecting human face and characteristics information extraction method or the method for detecting human face developed in the future and feature extracting method, Can be applied in target object information generation method according to embodiments of the present invention, and also should include the guarantor in the present invention In the range of shield.
In step s303, digitization is carried out to the characteristic information of described destination object, to form described destination object Format data.
By by the characteristic information data of described destination object, and form the format data of described destination object, make Obtain the transmission of succeeding target characteristics of objects information and the resource of occupancy required for identification greatly reduces, transmit beneficial to improving and know Other efficiency, and save transmission and computing resource.
Exemplarily, first digitization is carried out to the face characteristic information obtaining in step s302 taking facial image as a example Process, these information represented with data/numerical value, character, such as round face is expressed as 0, long face is expressed as 1 etc., and eye, nose, eyebrow, The positions such as mouth, shape, size etc. all can be converted into corresponding data, for example, be converted into coordinate, the number of expression shapes and sizes Value or character etc..
After by characteristic information data, then the characteristic information of these digitizations is formed with setting data format combination The data formatting, that is, form the format data of destination object.In the present embodiment, described data form adopts array data The characteristic information of these digitizations, that is, after by characteristic information data, is then arranged by form with setting data form To form array data, thus obtaining the format data of a facial image.Exemplarily, the digitization of described destination object Characteristic information includes data head, video camera information stamp, timestamp, address stamp, key point information, data tail.
In the present embodiment, the form of described array data is as shown in aforementioned table 1.Data is contained in this data form Head, video camera information stamp, timestamp, address stamp, face information, data tail etc., based on format data can obtain such as, take the photograph The data of camera model, image capturing time, place, image feature information etc., to be identified and to process.Due to array data Shared flow and/or bandwidth are significantly smaller than the data of picture or video format, so the data after array is to Netowrk tape Wide requirement substantially reduces, beneficial to the real-time of the quality improving data transfer.
It is understood that table 1 is only an exemplary array data form, the present invention can adopt extended formatting Array data.Additionally, array data is also only an example of data form of the present invention, the present invention can adopt suitable data Form, and it is not limited to above-mentioned array data, for example can adopt the data form of aggregate type, or can adopt self-defined Data form.
Specific embodiment, specifically can participate in the detailed description being directed to table 1 in preceding aim object identifying method embodiment, For sake of simplicity, will not be described here.
Exemplarily, target object information generation method according to embodiments of the present invention can have imageing sensor and Realize in the unit of processor or system.
Target object information generation method according to embodiments of the present invention can be deployed at target image collection terminal, example As in security protection application, the image acquisition end of gate control system can be deployed in；In financial application field, can be deployed in individual People's end, smart phone, panel computer, personal computer etc..
According to the target object information generation method of the present embodiment, by the characteristic information of destination object is formatted Process, to form the format data of destination object, so when spreading out of the format data of this destination object by network, Just the data volume of network transmission can be substantially reduced, save the network bandwidth and resource, improve the reality of destination object characteristic information transmission Shi Xing.
Fig. 4 is to compare a schematic flow according to data retrieval in the recongnition of objects method of the embodiment of the present invention Figure.With reference to Fig. 4, a kind of data retrieval comparison method in recongnition of objects method according to embodiments of the present invention is carried out Description.
As shown in figure 4, data retrieval according to embodiments of the present invention compares including:
In step 401, sequentially transfer the format data comparing object from data base.In the enforcement according to the present invention In example, in order to improve recognition efficiency, it is stored with data base and compares the format data of object, and in general, data Be stored with storehouse multiple or substantial amounts of comparison object.The type of destination object, e.g. facial image in abovementioned steps Or car plate, the then corresponding data base of type selecting according to destination object in this step, and sequentially adjust from this data base Take and transfer the format data comparing object one by one, compared with the format data with destination object, to determine target pair As if no whether similar to this if comparing object.
In step 402, the format data based on destination object and the format data currently comparing object provide phase Like degree evaluation of estimate.
Exemplarily it is however generally that the format data of destination object includes the characteristic information of multiple digitizations, such as face The characteristic information of the digitization of portion, eye, mouth, nose etc., compares for each digitization characteristic information in this step, i.e. example As relatively corresponding digitization characteristic information numerical value whether close to or character implication is identical or type is identical, be then based on Comparison result is evaluated to each digitization characteristic information respectively, and the evaluation being finally based on each digitization characteristic information is given Destination object and the similarity evaluation currently comparing object.
Further, when providing destination object with the similarity evaluation value currently comparing object, each digitization feature The weight of information can be the same or different.For example, in one example, concrete key character information, such as freckle, knife are set Scar, oxeye are important characteristic information, when destination object with compare when object differs larger in the comparing of this feature information then Will be considered that destination object and compare object dissmilarity, provide relatively low similarity evaluation, and without other characteristic informations of consideration be No close, when destination object with compare object in the comparing of this feature information close to or consistent when just consider other characteristic informations Whether close or consistent, and provide corresponding similarity evaluation.In another example, then set the weight phase of each characteristic information With weight shared by the evaluation of each characteristic information when providing similarity evaluation is identical or close, such as respectively account for 1/n, and n is characterized Information content.
It is likewise possible to participate in the detailed description of related embodiment in preceding aim object identifying method, for sake of simplicity, This repeats no more.
It should be appreciated that the above is only the schematic example providing similarity evaluation, the invention is not restricted to this, other are existing The similarity evaluation of similarity evaluation or later new settings is encompassed by within the scope of the present invention.
In step 403, judge destination object with whether the similarity currently comparing object is more than given threshold.
Exemplarily, for example given threshold is 90%, just can sentence when only providing 90% and above similarity evaluation Set the goal object to currently to compare object similar.Certainly, given threshold can flexibly be arranged as needed, and in different phase Can have different numerical value, rather than one is set and can not change.
If destination object is more than or equal to given threshold with the similarity currently comparing object, enter in step s204, Recall the current identity information comparing object.This identity information for example can include name, sex, age, height, body weight, family Nationality address etc..
If destination object is less than given threshold with the similarity currently comparing object, enters in step s404, judge Destination object is with currently to compare object dissimilar, and goes successively in step s405.
In step s405, judge whether current comparison object is that last compares object, terminate if in retrieval, Enter the identification starting next destination object in step s201, continue next comparison in step s401 if not then entering The comparison of object.
Exemplarily, the data retrieval comparison method according to the present embodiment can have setting of memorizer and processor Realize in standby, device or system.Data retrieval comparison method according to embodiments of the present invention can be deployed in facial image to be known At other identification end, control process end of gate control system of the master control end of such as preventing road monitoring system, Companies House etc. etc..
According to the data retrieval comparison method of the present embodiment, to be determined whether there is by directly carrying out data retrieval comparison Consistent comparison object, and directly carries out compared with image retrieval compares, reducing complexity and the treating capacity of identification business, improving Recognition efficiency is it is ensured that the real-time of recongnition of objects.
It is understood that described above is only the example that data retrieval compares, the data retrieval ratio of the present invention To method not limited to this, such as, except including said one or several step, other steps can also be included, such as in an example In, when obtaining the comparison object that multiple similarities are more than given threshold, then directly select similarity highest and compare object work For other to result, or the comparison object that all similarities are more than given threshold all selects comparison result, and the latter is to all phases It is more than the comparison object of given threshold according to new comparison condition like degree, again compared, these are all covered the present invention's In the range of.
Fig. 5 is to compare another schematic flow according to data retrieval in the recongnition of objects method of the embodiment of the present invention Figure.With reference to Fig. 5, data retrieval comparison method another kind of in recongnition of objects method according to embodiments of the present invention is entered Row description.
As shown in figure 5, the data retrieval comparison method of the present embodiment includes:
First, in step 501, by parallel for multiple digitization characteristic informations of destination object with the corresponding number in data base Compare according to changing characteristic information, such as in step 510, by destination object the first digitization characteristic information and institute in data base The the first digitization characteristic information comparing object is had to compare；In step 512, destination object the second digitization feature is believed Compare by all the second digitization characteristic informations comparing object with data base for breath；... in step 51n, the n-th data Change characteristic information, all the n-th digitization features comparing object are believed with data base by destination object the n-th digitization characteristic information Breath is compared.
In step 502, determine each digitization characteristic information of destination object whether there is in data base identical and The digitization characteristic information being close.
For example, in step 520, whether determine in data base in the first digitization characteristic information comparing object and target Object the first digitization characteristic information is identical or close；In step 521, determine in data base and whether comparing the second of object Digitization characteristic information is identical or close with destination object the second digitization characteristic information；In step 52n, determine in data base Whether identical or close with destination object the n-th digitization characteristic information in the n-th digitization characteristic information comparing object.
In step 503, for each digitization characteristic information, all close comparison objects are extracted.For example, in step In 530, extract and identical or close with the first digitization characteristic information of destination object all compare object；In step 531, Extract and identical or close with the second digitization characteristic information of destination object all compare object；In step 53n, extract with The identical or close all comparison objects of the digitization characteristic information of destination object.
In step 504, judge whether that the same or like quantity of digitization characteristic information is more than the ratio of predetermined value To object.
Exemplarily, can be by judging whether to compare the number of times that object occurs in step 530～step 530n To determine whether there is the comparison object that the same or like quantity of digitization characteristic information is more than predetermined value more than set point number. When there is the comparison object that the same or like quantity of digitization characteristic information is more than predetermined value, then enter in step s204, Recall the identity information of this comparison object, such as name, sex, age, height, native place etc..If there is no digitization feature The same or like quantity of information is more than the comparison object of predetermined value, then enter in step 201, continue the knowledge of next destination object Not.
Further, similar with previous embodiment, in the present embodiment, the weight of each digitization characteristic information can phase Together can also be different.When the weighted of each digitization characteristic information, for example when existing or setting key character information, then When this digitization characteristic information does not find same or close comparison object in step 502, then can directly tie Bundle retrieval compares, subsequently into the identification in step 201, continuing next destination object.
Exemplarily, the data retrieval comparison method according to the present embodiment can have setting of memorizer and processor Realize in standby, device or system.Data retrieval comparison method according to embodiments of the present invention can be deployed in facial image to be known At other identification end, control process end of gate control system of the master control end of such as preventing road monitoring system, Companies House etc. etc..
According to the data retrieval comparison method of the present embodiment, parallel search comparison is carried out to each digitization characteristic information, Further increase retrieval comparison efficiency, so that recognition efficiency further improves.
Fig. 6 is the schematic block diagram of the target object information signal generating unit according to the embodiment of the present invention.With reference to Fig. 6 is described to target object information signal generating unit according to embodiments of the present invention, and wherein, described target object information generates Unit is integrated in video camera.
As shown in fig. 6, target object information signal generating unit 600 according to embodiments of the present invention includes destination object detection carrying Delivery block 610, characteristic information extracting module 620 and format data generation module 630.
Destination object Detection and Extraction module 610 is used for detecting from the image of imageing sensor collection and extracting described target Object.Destination object Detection and Extraction module 610 can processor 102 Running storage device in electronic equipment as shown in Figure 1 In 104, the programmed instruction of storage to be realizing, and can execute target object information generation method according to embodiments of the present invention In step s301.
Characteristic information extracting module 620 is used for extracting the characteristic information of described destination object.That is, carry from destination object detection The characteristic information of destination object is extracted further in the destination object that delivery block 610 extracts.So-called characteristic information, refer to is permissible The information for identification object or destination object and other objects distinguished.Characteristic information extracting module 620 can be by In processor 102 Running storage device 104 in electronic equipment shown in Fig. 1, the programmed instruction of storage is realizing and permissible Execute step s302 in target object information generation method according to embodiments of the present invention.
Format data generation module 630 is used for carrying out digitization to the characteristic information of described destination object, to form bag Format data containing described destination object characteristic information.Format data generation module 630 can electronics as shown in Figure 1 set In processor 102 Running storage device 104 in standby, the programmed instruction of storage to be realizing, and can execute real according to the present invention Apply step s303 in the target object information generation method of example.
Described destination object Detection and Extraction module 610 includes coordinate measurement submodule (not shown) and image takes son Module (not shown).Coordinate measurement submodule, detects described target in the image that gathers from described image sensor The coordinate data of object position；Image takes submodule, for the coordinate data according to described destination object, from described figure As taking the image only including described destination object in the image of sensor acquisition.
Exemplarily, the format data of described destination object is the digitization characteristic information of described destination object to set The array data of data form arrangement form.
Exemplarily, the digitization characteristic information of described destination object include data head, video camera information stamp, timestamp, Address stamp, key point information, data tail.
According to the target object information signal generating unit of the present invention, by the characteristic information of destination object is formatted place Reason, to form the format data of destination object, so when the format data of this destination object is passed through network transmission, just The data volume of network transmission can be substantially reduced, save the network bandwidth and resource, improve the real-time of destination object characteristic information transmission Property.
Fig. 7 is the schematic block diagram of the recongnition of objects unit according to the embodiment of the present invention.With reference to Fig. 7 pair Recongnition of objects unit according to embodiments of the present invention is described, and exemplarily, recongnition of objects unit is integrated in clothes Business device end (or high in the clouds).
As shown in fig. 7, recongnition of objects unit 700 according to embodiments of the present invention includes information receiving module 710, inspection Rope comparing module 720 and judging treatmenting module 730.
Information receiving module 710 is used for receiving the format data of described destination object by network.Described destination object Format data be the digitization characteristic information comprising described destination object being generated by video camera.Information receiving module 710 Can the communication interface 108 in the processor 102 in electronic equipment as shown in Figure 1 realize.When being received by communication interface 108 To after the format data of destination object, can be deposited in memorizer 104 it is also possible to be transmitted directly to out processor 102 Processed.
Retrieval comparing module 720 is used for receiving the format data of described destination object from information receiving module 710, and will The format data comparing object in the format data of described destination object and data base carries out data retrieval and compares, with true Object is compared with the presence or absence of consistent with described destination object in fixed described data base.Retrieval comparing module 720 can be by Fig. 1 institute In processor 102 Running storage device 104 in the electronic equipment showing, the programmed instruction of storage to be realizing, and can execute root According to the step 202 in the target object information generation method of the embodiment of the present invention, 203, step s401～s405, and step S501 is to step 504.
In retrieval comparing module 720, processing module 730 determines that in described data base, presence is consistent with described destination object When comparing object, transfer the identity information that compare object corresponding to consistent with described destination object；In retrieval comparing module 720 Determine do not exist consistent with described destination object when comparing object in described data base, instruction proceeds next destination object Identification.Processing module 730 can storage in processor 102 Running storage device 104 in electronic equipment as shown in Figure 1 Programmed instruction is realizing, and can execute the step 204 in target object information generation method according to embodiments of the present invention With 205.
Exemplarily, the format data of the described destination object that described information receiver module 710 receives is by described shooting Machine generates.
Exemplarily, the format data of described destination object is the digitization characteristic information of described destination object to set The array data of data form arrangement form.
Exemplarily, the digitization characteristic information of described destination object include data head, video camera information stamp, timestamp, Address stamp, key point information, data tail.
Exemplarily, digitization in the format data to described destination object for the described retrieval comparing module 720 is special Reference breath carries out parallel search comparison.
Exemplarily, described retrieval comparing module 720 is carrying out being directed to each described comparison when described data retrieval compares Object provides similarity evaluation value.
Exemplarily, described retrieval comparing module 720 is carrying out being based on described destination object when described data retrieval compares Format data in each digitization characteristic information and data base in compare format data the and described target of object The similarity of each digitization characteristic information corresponding digitization characteristic information in the format data of object is directed to each institute State comparison object and provide similarity evaluation value.
Exemplarily, described retrieval comparing module 720 determines institute based on each described Similarity value evaluation comparing object State in data base and compare object with the presence or absence of consistent with described destination object, wherein, described retrieval comparing module is at least being deposited When a similarity evaluation value is more than the comparison object of given threshold, judge exist and described destination object in described data base Consistent comparison object, and select the corresponding object that compares of maximum similarity evaluation of estimate to be compare consistent with described destination object Object；When not depositing the comparison object that similarity evaluation is more than given threshold, then judge not exist in described data base with described The consistent comparison object of destination object.
Recongnition of objects unit according to embodiments of the present invention, by receiving the format data of described destination object, And the formatted data comparing object in the format data based on this destination object and data base carries out data retrieval and compares Determine whether there is consistent comparison object, and directly carry out compared with image retrieval compares, due to only needing transmission formatting data, Thus reduce the data volume of network transmission, and due to entering line retrieval using format data and compare to reduce identification business Complexity and treating capacity, improve recognition efficiency it is ensured that the real-time of recongnition of objects.
Fig. 8 is the schematic block diagram of the recongnition of objects system according to the embodiment of the present invention.With reference to Fig. 8 pair Recongnition of objects system according to embodiments of the present invention is described.
As shown in figure 8, recongnition of objects system 800 according to embodiments of the present invention includes imageing sensor 810, target Object information signal generating unit 600, communication unit 820, recongnition of objects unit 700 and transmitting element 830.
Imageing sensor 810 can be using various types of imageing sensors at present, and it can be black and white sensor or coloured silk Colour sensor, and imageing sensor can gather picture and can also gather video.Certainly, gather number with imageing sensor 810 According to process can come the programmed instruction of storage in processor 102 Running storage device 104 in electronic equipment as shown in Figure 1 Realize it is also possible to corresponding with special isp (image processor) or dsp (digital signal processor) instruct or firmware reality Existing, to obtain picture or video data.
Target object information signal generating unit 600 is used for the extracting target from images pair from described image sensor 810 collection As, and the characteristic information of destination object, and digitization and formatting are carried out to described destination object characteristic information, to be formed State the format data of destination object.The 26S Proteasome Structure and Function of target object information signal generating unit 600 and the description class with reference to Fig. 6 Seemingly, will not be described here.
The form of the destination object for being generated target object information signal generating unit 600 by network for the communication unit 820 Change data is activation to recongnition of objects unit 700, and in imageing sensor 810, target object information signal generating unit 600 and mesh Mark object identification unit 700 transmission control command, such as recongnition of objects unit 700 can be as needed to imageing sensor 810 transmitting control commands, indicate imageing sensor 810 adjustment position.Communication unit 820 can electronic equipment as shown in Figure 1 In processor 102 in communication interface 108 realize, it can be by the Internet, dedicated network or cable network, wireless network Network, Cellular Networks or internal system communication line are realized.
Recongnition of objects unit 700 is used for receiving the lattice of the destination object of target object information signal generating unit 600 generation Formula data, and the format data comparing object in the format data of described destination object and data base is carried out data inspection Rope compares, and compares object to determine in described data base with the presence or absence of consistent with described destination object, and carries out respective handling. The 26S Proteasome Structure and Function of recongnition of objects unit 700 is similar with the description with reference to Fig. 7, will not be described here.
Transmitting element 830 is used for recongnition of objects unit 700 identifies the identity of successful destination object by network Information sends to superior system to carry out subsequent treatment.Transmitting element 830 can processor in electronic equipment as shown in Figure 1 Communication interface 108 in 102 is realized, its can by the Internet, dedicated network or cable network, wireless network, Cellular Networks or Internal system communication line is realized.
Exemplarily, recongnition of objects system according to embodiments of the present invention can be implemented with imageing sensor, The unit of memorizer, communication interface/unit and processor or system.
Recongnition of objects system according to embodiments of the present invention can be deployed in face identification system, for example, in peace Anti- application, can be deployed in gate control system or preventing road monitoring system；In financial application field, can be with bank, exchange etc. Deng identity authorization system in.
Alternatively, recongnition of objects system according to embodiments of the present invention can also be deployed in server end with being distributed (or high in the clouds) and front-end image collection terminal.For example, in financial application field, imageing sensor and target object information can be given birth to Unit, communication unit is become to be deployed at personal terminal, smart phone, panel computer, personal computer etc., by destination object Recognition unit, transmitting element are deployed in server end (or high in the clouds), generate the formatting number of destination object at personal terminal According to, and sent to server end (or high in the clouds) by communication unit.Server end (or high in the clouds) is according to the destination object receiving Format data carries out data retrieval comparison, to identify destination object.Again for example, in preventing road monitoring system, can be by image Sensor and target object information signal generating unit, communication unit are deployed in remote monitoring device, by recongnition of objects unit, Transmitting element is deployed in server end (or high in the clouds), and remote monitoring device generates the format data of destination object, and leads to Cross communication unit to send to server end (or high in the clouds).Server end (or high in the clouds) is according to the formatting number of the destination object receiving According to carrying out data retrieval comparison, to identify destination object.
Further, when using distributed deployment, fore device and back-end server can be one-one relationship or many One-one relationship, than for n:1, wherein n is the natural number more than or equal to 1 for the quantity of for example described fore device and back-end server. Further, the format data of described destination object is the digitization characteristic information of described destination object with setting data form The array data of arrangement form.
Further, the digitization characteristic information of described destination object include data head, video camera information stamp, timestamp, Address stamp, key point information, data tail.
Further, distributed object object recognition system according to embodiments of the present invention, on the one hand, by destination object The image of information acquisition device collection destination object, and extract the characteristic information of destination object, and it is based on this feature information shape Become the format data of destination object, and the format data of this destination object is sent to by recongnition of objects list by network Unit, so due to only needing to send format data, therefore between send video or picture compared with greatly reduce network transmission Data volume, improve the real-time of network transmission, and saved the network bandwidth and resource；On the other hand known by destination object The format data based on this destination object for the other unit carries out data retrieval ratio with the formatted data comparing object in data base To determining whether there is consistent comparison object, so with directly carry out compared with image retrieval compares, due to format data Retrieval similar to conventional characters data retrieval, operand substantially reduces, thus reduces complexity and the place of identification business Reason amount, improves recognition efficiency it is ensured that the real-time of recongnition of objects, is suitable to meet extensive recongnition of objects business Require.
Additionally, according to embodiments of the present invention, additionally providing a kind of storage medium, storing program on said storage Instruction, gives birth to for executing the target object information of the embodiment of the present invention when described program instruction is run by computer or processor One-tenth method, the corresponding steps of recongnition of objects method, and for realizing target object information according to embodiments of the present invention Corresponding module in signal generating unit, recongnition of objects unit, recongnition of objects system.Described storage medium for example can wrap Include the storage card of the smart phone, memory unit of panel computer, the hard disk of personal computer, read only memory (rom), erasable Programmable read only memory (eprom), portable compact disc read only memory (cd-rom), usb memorizer or above-mentioned storage The combination in any of medium.Described computer-readable recording medium can be any of one or more computer-readable recording mediums Combination, such as one computer-readable recording medium comprises the computer-readable program generation generating for target object information Code, another computer-readable recording medium comprises the computer-readable program code for target object information identification.
In one embodiment, described computer program instructions can be realized when being run by computer according to the present invention in fact Apply the target object information signal generating unit of example and each functional module of recongnition of objects unit, and/or can execute Recongnition of objects method according to embodiments of the present invention, target object information generation method.
In one embodiment, described computer program instructions execute following steps when being run by computer: from image Detect in the image of sensor acquisition and extract destination object；Extract the characteristic information of described destination object；To described target pair The characteristic information of elephant carries out digitization and formatting, to form the format data comprising described destination object characteristic information.Show Example property ground, abovementioned steps are all executed by video camera.
Described computer program instructions also execute following steps when being run by computer: receive the lattice of described destination object Formula data, the format data of described destination object is the digitization feature comprising described destination object being generated by video camera Information；Carry out data retrieval ratio by compare the format data of object in the format data of described destination object and data base Right；Determined in described data base based on retrieval comparison result and compare object with the presence or absence of consistent with described destination object, if Exist in described data base consistent with described destination object compare object, then transfer consistent with described destination object compare right As corresponding identity information；Conversely, then proceeding the identification of next destination object.Exemplarily, abovementioned steps are by servicing Device end or high in the clouds execution.
Each module in recongnition of objects system according to embodiments of the present invention can be by according to embodiments of the present invention The processor of the electronic equipment of recongnition of objects run the computer program instructions that store in memory to realize, or The computer instruction that can store in the computer-readable recording medium of computer program according to embodiments of the present invention Realize when being run by computer.
Recongnition of objects method according to embodiments of the present invention and unit, information generating method and unit, destination object Identifying system and storage medium, by the characteristic information of destination object is carried out digitization and formatting, from regardless of whether transmit Or the resource taking required for identifying all substantially reduces, and improves recognition efficiency and real-time, it is suitable to meet extensive identification The requirement of business.
Although here by reference to Description of Drawings example embodiment it should be understood that above-mentioned example embodiment is merely exemplary , and be not intended to limit the scope of the invention to this.Those of ordinary skill in the art can carry out various changes wherein And modification, it is made without departing from the scope of the present invention and spirit.All such changes and modifications are intended to be included in claims Within required the scope of the present invention.
Those of ordinary skill in the art are it is to be appreciated that combine the list of each example of the embodiments described herein description Unit and algorithm steps, being capable of being implemented in combination in electronic hardware or computer software and electronic hardware.These functions are actually To be executed with hardware or software mode, the application-specific depending on technical scheme and design constraint.Professional and technical personnel Each specific application can be used different methods to realize described function, but this realization is it is not considered that exceed The scope of the present invention.
It should be understood that disclosed equipment and method in several embodiments provided herein, can be passed through it Its mode is realized.For example, apparatus embodiments described above are only schematically, for example, the division of described unit, and only It is only a kind of division of logic function, actual can have other dividing mode when realizing, and for example multiple units or assembly can be tied Close or be desirably integrated into another equipment, or some features can be ignored, or do not execute.
In description mentioned herein, illustrate a large amount of details.It is to be appreciated, however, that the enforcement of the present invention Example can be put into practice in the case of not having these details.In some instances, known method, structure are not been shown in detail And technology, so as not to obscure the understanding of this description.
Similarly it will be appreciated that in order to simplify the present invention and help understand one or more of each inventive aspect, In description to the exemplary embodiment of the present invention, each feature of the present invention be sometimes grouped together into single embodiment, figure, Or in descriptions thereof.However, not should by this present invention method be construed to reflect an intention that i.e. claimed Application claims than the feature being expressly recited in each claim more features.More precisely, as accordingly As claims are reflected, its inventive point is can be with all features of embodiment single disclosed in certain Feature is solving corresponding technical problem.Therefore, it then follows claims of specific embodiment are thus expressly incorporated in this tool Body embodiment, wherein each claim itself is as the separate embodiments of the present invention.
It will be understood to those skilled in the art that in addition to mutually exclusive between feature, any combinations pair can be adopted All features disclosed in this specification (including adjoint claim, summary and accompanying drawing) and so disclosed any method Or all processes of equipment or unit are combined.Unless expressly stated otherwise, (including adjoint right will for this specification Ask, make a summary and accompanying drawing) disclosed in each feature can be replaced by the alternative features providing identical, equivalent or similar purpose.
Although additionally, it will be appreciated by those of skill in the art that some embodiments described herein include other embodiments In included some features rather than further feature, but the combination of the feature of different embodiment means to be in the present invention's Within the scope of and form different embodiments.For example, in detail in the claims, embodiment required for protection one of arbitrarily Can in any combination mode using.
The all parts embodiment of the present invention can be realized with hardware, or to run on one or more processor Software module realize, or with combinations thereof realize.It will be understood by those of skill in the art that can use in practice Microprocessor or digital signal processor (dsp) are realizing some moulds in article analytical equipment according to embodiments of the present invention The some or all functions of block.The present invention is also implemented as a part for executing method as described herein or complete The program of device (for example, computer program and computer program) in portion.Such program realizing the present invention can store On a computer-readable medium, or can have the form of one or more signal.Such signal can be from the Internet Download on website and obtain, or provide on carrier signal, or provided with any other form.
It should be noted that above-described embodiment the present invention will be described rather than limits the invention, and ability Field technique personnel can design alternative embodiment without departing from the scope of the appended claims.In the claims, Any reference markss between bracket should not be configured to limitations on claims.Word "comprising" does not exclude the presence of not Element listed in the claims or step.Word "a" or "an" before element does not exclude the presence of multiple such Element.The present invention can come real by means of the hardware including some different elements and by means of properly programmed computer Existing.If in the unit claim listing equipment for drying, several in these devices can be by same hardware branch To embody.The use of word first, second, and third does not indicate that any order.These words can be explained and run after fame Claim.
The above, the only specific embodiment of the present invention or the explanation to specific embodiment, the protection of the present invention Scope is not limited thereto, any those familiar with the art the invention discloses technical scope in, can be easily Expect change or replacement, all should be included within the scope of the present invention.Protection scope of the present invention should be with claim Protection domain is defined.

Claims (28)

1. a kind of recongnition of objects method is it is characterised in that include:
Receive the format data of described destination object, the format data of described destination object is by comprising that video camera generates The digitization characteristic information of described destination object；
The format data comparing object in the format data of described destination object and data base is carried out data retrieval compare；
Determined in described data base based on retrieval comparison result and compare object with the presence or absence of consistent with described destination object,
If what in described data base, presence was consistent with described destination object compares object, transfer consistent with described destination object Compare object corresponding to identity information.
2. recongnition of objects method according to claim 1 is it is characterised in that the format data of described destination object Obtained by described video camera execution following steps:
Collection includes the image of destination object；
Destination object is extracted from described image；
The characteristic information of destination object is extracted from described destination object；
By the characteristic information dataization of described destination object and formatting to form the format data of described destination object.
3. recongnition of objects method according to claim 1 is it is characterised in that the format data of described destination object For described destination object digitization characteristic information with the array data of setting data form arrangement form.
4. recongnition of objects method according to claim 3 is it is characterised in that the digitization feature of described destination object Information includes data head, video camera information stamp, timestamp, address stamp, key point information, data tail.
5. recongnition of objects method according to claim 1 is it is characterised in that when carrying out described data retrieval and comparing Parallel search comparison is carried out to each digitization characteristic information in the format data of described destination object.
6. recongnition of objects method according to claim 1 is it is characterised in that when carrying out described data retrieval and comparing Provide similarity evaluation value for each described comparison object.
7. recongnition of objects method according to claim 5 is it is characterised in that when carrying out described data retrieval and comparing The formatting of object is compared in each digitization characteristic information in format data based on described destination object and data base Each digitization characteristic information corresponding digitization characteristic information in the format data with described destination object of data Similarity is directed to each described comparison object and provides similarity evaluation value.
8. the recongnition of objects method according to claim 6 or 7 is it is characterised in that be based on each described comparison object Similarity evaluation value determine whether there is in described data base consistent with described destination object compare object,
Wherein, when the similarity evaluation value that at least there is a comparison object is more than given threshold, judge in described data base Exist consistent with described destination object compare object, and select maximum similarity evaluation of estimate corresponding comparison object be with described The consistent comparison object of destination object；Conversely, then judging not existing compare consistent with described destination object in described data base Object.
9. a kind of target object information generation method is it is characterised in that include:
Detect from the image of imageing sensor collection and extract destination object；
Extract the characteristic information of described destination object；
Digitization and formatting are carried out to the characteristic information of described destination object, comprises described destination object characteristic information to be formed Format data；
Wherein, abovementioned steps are all executed by video camera.
10. target object information generation method according to claim 9 is it is characterised in that described adopt from imageing sensor Detect and extract destination object in the image of collection and include:
The coordinate data of described destination object position is detected from the image of described image sensor collection；
According to the coordinate data of described destination object, take from the image of described image sensor collection and only include described target The image of object.
11. target object information generation methods according to claim 9 or 10 are it is characterised in that described destination object Format data is the digitization characteristic information of described destination object with the array data of setting data form arrangement form.
12. target object information generation methods according to claim 9 or 10 are it is characterised in that described destination object Digitization characteristic information includes data head, video camera information stamp, timestamp, address stamp, key point information, data tail.
A kind of 13. target object information signal generating units are it is characterised in that described target object information signal generating unit is integrated in shooting In machine, and described target object information signal generating unit includes:
Destination object Detection and Extraction module, described destination object Detection and Extraction module is used for from the image of imageing sensor collection Detect and extract destination object；
Characteristic information extracting module, described characteristic information extracting module extracts the feature letter of destination object from described destination object Breath；
Format data generation module, described format data generation module is used for the characteristic information data of described destination object Change and format, to form the format data comprising described destination object characteristic information.
14. target object information signal generating units according to claim 13 are it is characterised in that the detection of described destination object carries Delivery block includes:
Coordinate measurement submodule, detects described destination object position in the image that gathers from described image sensor Coordinate data；
Image takes submodule, for the coordinate data according to described destination object, from the image of described image sensor collection In take the image only including described destination object.
15. target object information signal generating units according to claim 13 or 14 are it is characterised in that described destination object Format data is the digitization characteristic information of described destination object with the array data of setting data form arrangement form.
16. target object information signal generating units according to claim 13 or 14 are it is characterised in that described destination object Digitization characteristic information includes data head, video camera information stamp, timestamp, address stamp, key point information, data tail.
A kind of 17. recongnition of objects units are it is characterised in that include:
Information receiving module, described information receiver module is used for receiving the format data of destination object, described mesh by network The format data of mark object is the digitization characteristic information comprising described destination object being generated by video camera；
Retrieval comparing module, described retrieval comparing module is used for the ratio in the format data and data base of described destination object Data retrieval comparison is carried out to the format data of object, be whether there is and described destination object one with determining in described data base The comparison object causing；
Processing module, described processing module determines in described data base exist and described destination object in described retrieval comparing module During consistent comparison object, transfer the identity information that compare object corresponding to consistent with described destination object.
18. recongnition of objects units according to claim 17 are it is characterised in that the formatting number of described destination object According to the digitization characteristic information for described destination object with the array data of setting data form arrangement form.
19. recongnition of objects units according to claim 17 are it is characterised in that the digitization of described destination object is special Reference breath includes data head, video camera information stamp, timestamp, address stamp, key point information, data tail.
20. recongnition of objects units according to claim 17 are it is characterised in that described retrieval comparing module is to institute The digitization characteristic information stated in the format data of destination object carries out parallel search comparison.
21. recongnition of objects units according to claim 17 are it is characterised in that described retrieval comparing module is being carried out Similarity evaluation value is provided for each described comparison object when described data retrieval compares.
22. recongnition of objects units according to claim 20 are it is characterised in that described retrieval comparing module is being carried out Each digitization characteristic information in format data based on described destination object when described data retrieval compares and data base Each digitization characteristic information in the format data with described destination object of format data of middle comparison object is corresponding Digitization characteristic information similarity be directed to each described comparison object provide similarity evaluation value.
23. recongnition of objects units according to claim 21 or 22 are it is characterised in that described retrieval comparing module base Determine in described data base with the presence or absence of consistent with described destination object in each described Similarity value evaluation comparing object Compare object,
Wherein, described retrieval comparing module at least exist a similarity evaluation value be more than given threshold comparison object when, Judge to exist in described data base consistent with described destination object compare object, and selection maximum similarity evaluation of estimate is corresponding Comparing object is consistent with described destination object to compare object；Do not depositing the comparison object that similarity evaluation is more than given threshold When, then judge not exist in described data base and consistent with described destination object compare object.
A kind of 24. recongnition of objects systems are it is characterised in that include:
Imageing sensor, described image sensor is used for gathering image；
Target object information signal generating unit, described target object information signal generating unit is used for the figure from described image sensor collection Acquisition destination object in picture, and extract the characteristic information of institute's destination object, and and described destination object characteristic information is carried out Digitization and formatting, to form the format data of described destination object；
Recongnition of objects unit, described recongnition of objects unit is used for receiving the format data of described destination object, and The format data comparing object in the format data of described destination object and data base is carried out data retrieval compare, with true Compare object with the presence or absence of consistent with described destination object in fixed described data base, and carry out respective handling,
Wherein, comparing that the presence in determining described data base of described recongnition of objects unit is consistent with described destination object is right As when, transfer the identity information that compare object corresponding to consistent with described destination object；Do not deposit in determining described data base Consistent with described destination object when comparing object, instruction proceeds the identification of next destination object；
Communication unit, described communication unit is configured to the described destination object of described target object information signal generating unit generation Format data sends to described recongnition of objects unit.
25. recongnition of objects systems according to claim 24 are it is characterised in that described image sensor, target pair Image information signal generating unit and communication unit configure in far-end image collecting device, and described recongnition of objects unit configuration is in clothes Business device end.
26. recongnition of objects systems according to claim 25 it is characterised in that described far-end image collecting device with Than for n:1, wherein n is the natural number more than or equal to 1 to the quantity of described server end.
27. recongnition of objects systems according to claim 24 are it is characterised in that the formatting number of described destination object According to the digitization characteristic information for described destination object with the array data of setting data form arrangement form.
28. recongnition of objects systems according to any one in claim 24-27 are it is characterised in that described mesh The digitization characteristic information of mark object includes data head, video camera information stamp, timestamp, address stamp, key point information, data Tail.
CN201610798111.0A 2016-08-31 2016-08-31 Target object information generation method and unit identification method and unit and system Pending CN106372606A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610798111.0A CN106372606A (en) 2016-08-31 2016-08-31 Target object information generation method and unit identification method and unit and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610798111.0A CN106372606A (en) 2016-08-31 2016-08-31 Target object information generation method and unit identification method and unit and system

Publications (1)

Publication Number Publication Date
CN106372606A true CN106372606A (en) 2017-02-01

Family

ID=57899342

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610798111.0A Pending CN106372606A (en) 2016-08-31 2016-08-31 Target object information generation method and unit identification method and unit and system

Country Status (1)

Country Link
CN (1) CN106372606A (en)

Cited By (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106993155A (en) * 2017-03-09 2017-07-28 北京溢思德瑞智能科技研究院有限公司 A kind of video flowing multi-processor array
CN108038468A (en) * 2017-12-26 2018-05-15 北斗七星（重庆）物联网技术有限公司 A kind of security terminal based on recognition of face
CN108647629A (en) * 2018-05-08 2018-10-12 北京大米科技有限公司 Image path recognition methods and device in a kind of image recognition localization method, video
CN108881846A (en) * 2018-07-10 2018-11-23 深圳市易成自动驾驶技术有限公司 Information fusion method, device and computer readable storage medium
CN109003449A (en) * 2018-08-05 2018-12-14 于澎 Remote vehicle information managing and control system
CN109116129A (en) * 2017-06-26 2019-01-01 深圳回收宝科技有限公司 Endpoint detection methods, detection device, system and storage medium
CN109165620A (en) * 2018-09-04 2019-01-08 张晓飞 A kind of detection method of electric vehicle, system and terminal device
CN109325548A (en) * 2018-10-23 2019-02-12 北京旷视科技有限公司 Image processing method, device, electronic equipment and storage medium
CN109376743A (en) * 2018-09-28 2019-02-22 北京旷视科技有限公司 Image processing method, device, image recognition apparatus and storage medium
CN109447154A (en) * 2018-10-29 2019-03-08 网易（杭州）网络有限公司 Picture similarity detection method, device, medium and electronic equipment
CN109522853A (en) * 2018-11-22 2019-03-26 湖南众智君赢科技有限公司 Face datection and searching method towards monitor video
CN109639700A (en) * 2018-12-25 2019-04-16 深圳市天彦通信股份有限公司 Personal identification method, device, equipment, cloud server and storage medium
CN109889773A (en) * 2017-12-06 2019-06-14 中国移动通信集团四川有限公司 Method, apparatus, equipment and the medium of the monitoring of assessment of bids room personnel
CN109947992A (en) * 2019-03-05 2019-06-28 北京旷视科技有限公司 Retrieve management-control method, head end video processing method, control control method and device
CN110110110A (en) * 2018-02-02 2019-08-09 杭州海康威视数字技术股份有限公司 One kind is to scheme to search drawing method, device, electronic equipment and storage medium
CN110147759A (en) * 2019-05-17 2019-08-20 北京迈格威科技有限公司 Target identification method, system, target identification management method and storage medium
CN110348315A (en) * 2019-06-14 2019-10-18 深圳英飞拓科技股份有限公司 Dynamic updates method and device, the face snap system in face characteristic bottom library
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
CN111291682A (en) * 2020-02-07 2020-06-16 浙江大华技术股份有限公司 Method and device for determining target object, storage medium and electronic device
CN112351402A (en) * 2020-11-07 2021-02-09 杭州集联科技有限公司 On-line detection system for MAC data acquisition unit

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102543125A (en) * 2012-01-17 2012-07-04 中兴通讯股份有限公司 Method and system for providing value-added service based on radio frequency identification technique
CN103268549A (en) * 2013-04-24 2013-08-28 徐明亮 Mobile payment verification system based on facial features
CN103514751A (en) * 2012-06-19 2014-01-15 广州市捷众科贸有限公司 Vehicle information identification apparatus
CN104463123A (en) * 2014-12-11 2015-03-25 南威软件股份有限公司 B/S-based face recognition method and system
CN105187217A (en) * 2015-09-16 2015-12-23 成都三零凯天通信实业有限公司 Dual-factor identity authentication method and apparatus thereof
CN105894262A (en) * 2016-04-01 2016-08-24 袁艳荣 Customer authentication system based on image identification

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102543125A (en) * 2012-01-17 2012-07-04 中兴通讯股份有限公司 Method and system for providing value-added service based on radio frequency identification technique
CN103514751A (en) * 2012-06-19 2014-01-15 广州市捷众科贸有限公司 Vehicle information identification apparatus
CN103268549A (en) * 2013-04-24 2013-08-28 徐明亮 Mobile payment verification system based on facial features
CN104463123A (en) * 2014-12-11 2015-03-25 南威软件股份有限公司 B/S-based face recognition method and system
CN105187217A (en) * 2015-09-16 2015-12-23 成都三零凯天通信实业有限公司 Dual-factor identity authentication method and apparatus thereof
CN105894262A (en) * 2016-04-01 2016-08-24 袁艳荣 Customer authentication system based on image identification

Cited By (28)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106993155A (en) * 2017-03-09 2017-07-28 北京溢思德瑞智能科技研究院有限公司 A kind of video flowing multi-processor array
CN109116129B (en) * 2017-06-26 2021-02-23 深圳回收宝科技有限公司 Terminal detection method, detection device, system and storage medium
CN109116129A (en) * 2017-06-26 2019-01-01 深圳回收宝科技有限公司 Endpoint detection methods, detection device, system and storage medium
CN109889773A (en) * 2017-12-06 2019-06-14 中国移动通信集团四川有限公司 Method, apparatus, equipment and the medium of the monitoring of assessment of bids room personnel
CN108038468A (en) * 2017-12-26 2018-05-15 北斗七星（重庆）物联网技术有限公司 A kind of security terminal based on recognition of face
CN110110110A (en) * 2018-02-02 2019-08-09 杭州海康威视数字技术股份有限公司 One kind is to scheme to search drawing method, device, electronic equipment and storage medium
CN108647629A (en) * 2018-05-08 2018-10-12 北京大米科技有限公司 Image path recognition methods and device in a kind of image recognition localization method, video
CN108881846A (en) * 2018-07-10 2018-11-23 深圳市易成自动驾驶技术有限公司 Information fusion method, device and computer readable storage medium
CN108881846B (en) * 2018-07-10 2020-12-08 深圳市易成自动驾驶技术有限公司 Information fusion method and device and computer readable storage medium
CN109003449A (en) * 2018-08-05 2018-12-14 于澎 Remote vehicle information managing and control system
CN109165620A (en) * 2018-09-04 2019-01-08 张晓飞 A kind of detection method of electric vehicle, system and terminal device
CN109165620B (en) * 2018-09-04 2022-10-11 张晓飞 Detection method and system of electric vehicle and terminal equipment
CN109376743A (en) * 2018-09-28 2019-02-22 北京旷视科技有限公司 Image processing method, device, image recognition apparatus and storage medium
CN109325548A (en) * 2018-10-23 2019-02-12 北京旷视科技有限公司 Image processing method, device, electronic equipment and storage medium
CN109325548B (en) * 2018-10-23 2021-03-23 北京旷视科技有限公司 Image processing method, image processing device, electronic equipment and storage medium
CN109447154A (en) * 2018-10-29 2019-03-08 网易（杭州）网络有限公司 Picture similarity detection method, device, medium and electronic equipment
CN109522853B (en) * 2018-11-22 2019-11-19 湖南众智君赢科技有限公司 Face datection and searching method towards monitor video
CN109522853A (en) * 2018-11-22 2019-03-26 湖南众智君赢科技有限公司 Face datection and searching method towards monitor video
CN109639700A (en) * 2018-12-25 2019-04-16 深圳市天彦通信股份有限公司 Personal identification method, device, equipment, cloud server and storage medium
CN109947992A (en) * 2019-03-05 2019-06-28 北京旷视科技有限公司 Retrieve management-control method, head end video processing method, control control method and device
CN110147759A (en) * 2019-05-17 2019-08-20 北京迈格威科技有限公司 Target identification method, system, target identification management method and storage medium
CN110348315A (en) * 2019-06-14 2019-10-18 深圳英飞拓科技股份有限公司 Dynamic updates method and device, the face snap system in face characteristic bottom library
CN110348315B (en) * 2019-06-14 2021-10-19 深圳英飞拓科技股份有限公司 Method and device for dynamically updating human face feature base and human face snapshot 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
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
CN111291682A (en) * 2020-02-07 2020-06-16 浙江大华技术股份有限公司 Method and device for determining target object, storage medium and electronic device
CN112351402A (en) * 2020-11-07 2021-02-09 杭州集联科技有限公司 On-line detection system for MAC data acquisition unit
CN112351402B (en) * 2020-11-07 2022-07-22 杭州集联科技有限公司 Online detection system for MAC data acquisition unit

Similar Documents

Publication Publication Date Title
CN106372606A (en) Target object information generation method and unit identification method and unit and system
CN109697416B (en) Video data processing method and related device
CN111027378B (en) Pedestrian re-identification method, device, terminal and storage medium
CN110781784A (en) Face recognition method, device and equipment based on double-path attention mechanism
CN113365147B (en) Video editing method, device, equipment and storage medium based on music card point
CN107993191A (en) A kind of image processing method and device
CN109766755B (en) Face recognition method and related product
CN107220633A (en) A kind of intelligent mobile enforcement system and method
CN110866236B (en) Private picture display method, device, terminal and storage medium
TWI712980B (en) Claim information extraction method and device, and electronic equipment
CN112307886A (en) Pedestrian re-identification method and device
CN109800304A (en) Processing method, device, equipment and the medium of case notes
CN110532866A (en) Video data detection method, device, computer equipment and storage medium
CN106951866A (en) A kind of face authentication method and device
WO2021203882A1 (en) Attitude detection and video processing method and apparatus, and electronic device and storage medium
CN110442742A (en) Retrieve method and device, processor, electronic equipment and the storage medium of image
CN110647938B (en) Image processing method and related device
WO2022246989A1 (en) Data identification method and apparatus, and device and readable storage medium
CN107679563A (en) Image processing method and device, system, computer equipment
CN108171208A (en) Information acquisition method and device
CN110322472A (en) A kind of multi-object tracking method and terminal device
CN109446946A (en) A kind of multi-cam real-time detection method based on multithreading
CN108133197A (en) For generating the method and apparatus of information
CN207037676U (en) Human face image collecting device
CN114882314A (en) Model training method and related product, image processing method and related product

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
CB02 Change of applicant information

Address after: 100190 Beijing, Haidian District Academy of Sciences, South Road, No. 2, block A, No. 313

Applicant after: MEGVII INC.

Applicant after: Beijing maigewei Technology Co., Ltd.

Address before: 100190 Beijing, Haidian District Academy of Sciences, South Road, No. 2, block A, No. 313

Applicant before: MEGVII INC.

Applicant before: Beijing aperture Science and Technology Ltd.

CB02 Change of applicant information
RJ01 Rejection of invention patent application after publication

Application publication date: 20170201

RJ01 Rejection of invention patent application after publication