WO2020238232A1 - 图像识别方法、装置、设备和认证方法、装置、设备 - Google Patents

图像识别方法、装置、设备和认证方法、装置、设备 Download PDF

Info

Publication number
WO2020238232A1
WO2020238232A1 PCT/CN2020/070909 CN2020070909W WO2020238232A1 WO 2020238232 A1 WO2020238232 A1 WO 2020238232A1 CN 2020070909 W CN2020070909 W CN 2020070909W WO 2020238232 A1 WO2020238232 A1 WO 2020238232A1
Authority
WO
WIPO (PCT)
Prior art keywords
target object
target
target image
image
identification mark
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.)
Ceased
Application number
PCT/CN2020/070909
Other languages
English (en)
French (fr)
Inventor
徐崴
陈继东
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.)
Advanced New Technologies Co Ltd
Original Assignee
Advanced New Technologies Co 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 Advanced New Technologies Co Ltd filed Critical Advanced New Technologies Co Ltd
Priority to US16/805,530 priority Critical patent/US10796199B1/en
Publication of WO2020238232A1 publication Critical patent/WO2020238232A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; 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
    • G06F18/211—Selection of the most significant subset of features
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
    • G06V10/225—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition based on a marking or identifier characterising the area

Definitions

  • the embodiments of this specification relate to the field of computer technology, and in particular to an image recognition method, device, device, and authentication method, device, and device.
  • forged electronic documents refer to the act of tampering with information on the obtained electronic document images after taking photos and scanning the physical documents, such as using PS( Photoshop) software tampering
  • forged electronic documents have the characteristics of easy acquisition of forged targets, low cost of forgery, easy acquisition of forgery tools, and realistic effects, which makes forged electronic documents often appear in the process of remote authentication of documents, which is safe for remote businesses. Brings a great threat.
  • the embodiments of this specification provide an image recognition method, device, device, and authentication method, device, and device, which are used for image category recognition.
  • the embodiment of this specification provides an image recognition method, including:
  • the attribute of the target object corresponding to the target image is determined according to the identification feature according to the preset mapping relationship, and the mapping relationship includes the corresponding relationship between the identification feature and the attribute of the target object.
  • the embodiment of this specification provides an authentication method, including:
  • the mapping relationship includes the corresponding relationship between the identification feature and the attribute of the target object
  • the target object is authenticated.
  • the embodiment of this specification provides an image recognition device, including:
  • An acquiring unit acquiring a target image of the target object, wherein, before acquiring the target image, an identification mark is mapped on the target object, and the identification mark is used to form a corresponding identification feature in the target image;
  • the recognition unit determines the attribute of the target object corresponding to the target image according to the preset mapping relationship according to the recognition feature, and the mapping relationship includes the corresponding relationship between the recognition feature and the attribute of the target object.
  • the embodiment of this specification provides an authentication device, including:
  • An acquiring unit acquiring a first target image of the target object, where before acquiring the first target image, the target object is mapped with an identification mark, and the identification mark is used to form a corresponding identification feature in the first target image;
  • the recognition unit determines the attributes of the target object corresponding to the first target image according to the preset mapping relationship according to the recognition feature, and the mapping relationship includes the corresponding relationship between the recognition feature and the attribute of the target object;
  • the authentication unit authenticates the target object according to the determined result.
  • the embodiment of the specification provides an electronic device, including: at least one processor and a memory, the memory stores a program and is configured to execute the following steps by the at least one processor:
  • the attribute of the target object corresponding to the target image is determined according to the identification feature according to the preset mapping relationship, and the mapping relationship includes the corresponding relationship between the identification feature and the attribute of the target object.
  • the embodiment of the specification provides an electronic device, including: at least one processor and a memory, the memory stores a program and is configured to execute the following steps by the at least one processor:
  • mapping relationship includes the corresponding relationship between the identification feature and the attribute of the target object
  • the target object is authenticated.
  • the target object is mapped with an identification mark, and the identification mark is used to form a corresponding identification feature in the target image, thereby using the difference between the identification mark on the target object,
  • the identification features that can be used for identification are formed on the target image, the identification features based on differences are actively added to the identification process, and the attributes of the target object corresponding to the target image are determined according to the identification features according to the preset mapping relationship.
  • the mapping relationship includes recognition The corresponding relationship between features and attributes of target objects effectively improves the accuracy of target image recognition.
  • FIG. 1 is a schematic diagram of the application of an image recognition method proposed in an embodiment of this specification.
  • Fig. 2 is a flowchart of an image recognition method proposed in an embodiment of this specification.
  • FIG. 3 is a schematic diagram of the collection of ID card images using a mobile phone in an image recognition method proposed in an embodiment of this specification.
  • Fig. 4 is a flowchart of an authentication method proposed in an embodiment of the specification.
  • Fig. 5 is a flowchart of an authentication method proposed in an embodiment of this specification.
  • Fig. 6 is a schematic structural diagram of an image recognition device proposed in an embodiment of the specification.
  • Fig. 7 is a schematic structural diagram of an authentication device proposed in an embodiment of this specification.
  • the electronic object to be authenticated for example, the image to be identified obtained from the certificate to be authenticated
  • the image to be identified obtained from the certificate to be authenticated can be identified by the following methods: obtain the image to be identified according to the object to be authenticated, extract the identification features from the image, and identify Features classify and recognize the picture, and then determine whether the object to be recognized comes from the original.
  • the object to be authenticated is an electronic piece forged according to the object to be authenticated, for example Forged electronic documents
  • the forged electronic parts must be presented in some physical way, for example, through screen display or print display, in order to obtain the image to be identified through a second copy, and then enter it into the authentication system Therefore, the image recognition method, device, device and authentication method, device, and device proposed in the embodiments of this specification make full use of the feature that forged electronic parts need to be displayed on the screen or printed to be used for authentication. Classes are identified based on the to-be-identified image obtained by the second re-shooting of the forged electronic parts, which effectively prevents the use of forged electronic documents in the authentication process.
  • the application schematic diagram of the image recognition process in the embodiment of this specification may be as shown in Fig. 1, including a target object 10 to be recognized, such as an ID card, an identification mark 20 mapped on the target object, such as a QR code, and according to the target object 10 and the target image 30 obtained by the identification mark 20.
  • a target object 10 to be recognized such as an ID card
  • an identification mark 20 mapped on the target object such as a QR code
  • the target image 30 is obtained according to the target object 10 to which the identification mark 20 is mapped, and the target image 30 has the identification feature formed according to the identification mark 20.
  • the preset mapping relationship is based on the identification feature
  • the target image 30 is recognized, and then the attributes of the target object corresponding to the target image 30 are determined.
  • mapping of the identification mark 20 and the collection, identification, and determination of the target image 30 may be performed by the same subject, or by different subjects.
  • the terminal maps the identification mark 20 on the identification object 10, acquires the target image 30, recognizes the target image 30, and determines the attribute of the target object corresponding to the target image 30.
  • the terminal maps the identification mark 20 on the identification object 10, acquires the target image 30, sends the target image 30 to the server for identification, and then sends the identification result to the terminal to determine the attributes of the target object corresponding to the target image 30.
  • the above-mentioned joint execution process can also be: mapping the identification mark 20 on the identification object 10 through the first terminal, acquiring the target image 30 through the second terminal, sending the target image 30 to the server for identification, and then sending the identification result to the second terminal.
  • the second terminal determines the attributes of the target object corresponding to the target image 30.
  • the terminal includes terminal devices that can run software, including but not limited to computers, tablet computers, mobile phones, bank client devices, and so on.
  • the overall idea of the image recognition method proposed in the embodiments of this specification is that the forged electronic parts must be displayed on the screen or printed display, so that the forged electronic parts and the original have different surface material characteristics.
  • the identification mark is mapped on the target object. Based on the reason that the surface material of the original and the fake electronic component is different, the identification mark will be different when displayed on the target object, and then the target image corresponding to the target object with the identification mark is obtained.
  • the identification feature can include the feature formed according to the difference of the identification mark on the target object, so as to actively add the identification feature to the identification process, and finally based on this difference in use Recognition features that have been actively increased, through the mapping relationship that has been trained, the attributes of the target object corresponding to the target image are determined.
  • the target image can be an image obtained from the original, or it can be obtained from a displayed fake electronic component. The image effectively improves the recognition accuracy of the target image.
  • the image recognition method proposed in the embodiment of this specification obtains the target image of the target object, wherein, before the target image is obtained, the target object is mapped with an identification mark, and the identification mark is used to form a corresponding identification feature in the target image, making full use of Based on the difference between the identification marks on different types of target objects, the identification features that can be recognized on the target image are formed based on this difference, which provides a wealth of identification information for subsequent recognition.
  • the identification marks are set by mapping to avoid Then, the attributes of the target object in the target image are recognized according to the recognition characteristics according to the preset mapping relationship, and the recognition judgment is made based on the recognition characteristics, which effectively improves the accuracy of recognition.
  • Such a recognition process actively adds easy-to-recognize features to the recognition process by mapping the recognition mark on the target object, and recognizes the picture based on the newly-added recognition feature, which improves both the accuracy and the recognition efficiency. Avoid interference of irrelevant features on the recognition process and recognition results.
  • Fig. 2 is a flowchart of an image recognition method proposed in an embodiment of this specification.
  • the image recognition method in the embodiment of this specification includes the following steps:
  • Step S201 Obtain a target image of the target object.
  • an identification mark is mapped on the target object, and the identification mark is used to form a corresponding identification feature in the target image.
  • the target object includes an object used to generate a target image.
  • the target object can be a document used to characterize the identity, such as an ID card, driving license, passport, etc.
  • the document may be the original document, or it may be forged based on the original document.
  • Electronic files for example, a copy of a certificate).
  • the target object can also be various forms of bank cards.
  • the bank card may be the original bank card or an electronic counterfeit based on the original. Pieces (for example, a remake of a bank card).
  • the remake is a reproduced piece that is physically presented again, for example, a remake displayed on a screen or a remake displayed on a print.
  • the identification mark includes a mark that is mapped on a target object and used to generate a feature for identification.
  • the identification mark may be a specific pattern.
  • the identification mark is mapped on the surface of the target object, which can be through the light source with the identification mark to emit light.
  • the light is reflected on the surface of the target object, and then on the target object The surface is mapped to the identification mark that can form the identification feature.
  • Such an identification mark mapping method can not only adjust the identification mark according to actual needs to meet the complex actual use requirements, but also avoid damage to the target object. At the same time, it can also set the identification mark on the target object simply, quickly and clearly. .
  • the identification feature includes the features that are formed in the target image and correspond to the identification mark and can be identified.
  • the target object used to obtain the target image has different surface materials. Therefore, when the identification mark is mapped to the target object, the identification mark will be different on different surfaces due to the different surface material.
  • the surface material is anti-counterfeiting coating.
  • the surface material is LCD screen. At this time, the identification mark will be on the surface of two different materials. Obvious differences are produced, and the identification features corresponding to the identification marks in the acquired target image will also retain these differences, and then the identification features can be used for identification based on these differences.
  • the identification mark in order to be able to identify more accurately, includes a pattern containing positioning features, and the positioning feature is used to locate the pattern containing the positioning feature.
  • the positioning feature contained in the pattern can quickly To locate the pattern, that is, to quickly locate the identification mark according to the positioning feature, and then identify it.
  • the aforementioned pattern containing positioning features may be a circular pattern, a symmetrical polygonal pattern, and the like.
  • the pattern containing the positioning feature is used as the identification mark, so that the identification mark can be located based on the positioning feature in the subsequent identification process, and the position of the identification feature is determined, thereby ensuring that the identification feature can be used for identification more accurately and quickly.
  • the acquired target image includes the target object and the identification feature corresponding to the identification mark.
  • the method of obtaining the target image may include shooting acquisition, for example, using a camera to shoot the target object and the identification mark, for example, taking a photo or video, and then obtaining the captured image.
  • shooting acquisition for example, using a camera to shoot the target object and the identification mark, for example, taking a photo or video
  • a photo or a picture in a video is used as the target image; it may also include scanning acquisition, for example, a scanner is used to scan the target object and the identification mark, and the scanned image is acquired as the target image.
  • the recognition features in the acquired target image include any one or more of color features, texture features, shape features, and spatial relationship features.
  • the mapping is The identification mark on it will produce grayscale, color difference, texture difference, contour deformation difference and relative position difference with ID card, etc. Based on these differences, the identification mark can be formed on the target image. Identify features.
  • the image recognition method in the embodiment of this specification can be used in a mobile terminal, and the mobile terminal includes a display unit and an image acquisition unit;
  • the steps of obtaining the target image of the target object include:
  • mapping the identification mark on the target object includes:
  • the image acquisition unit and the display unit are located on the same side of the target object.
  • a mobile terminal may include a mobile phone, a tablet, a computer, a bank client or other mobile devices capable of displaying images and performing image collection.
  • the display unit is used to present the identification mark and map the identification mark to the target object.
  • the identification mark may include a randomly generated identification mark, such as a round icon, a square icon, etc.; it may also be an identification mark containing preset specific information, such as trademarks, abbreviations, two-dimensional codes, barcodes, and serial numbers.
  • the display of the identification mark by the display unit may include displaying the same identification mark, for example, displaying the abbreviation of the bank as the identification mark; or displaying the identification mark according to a predetermined display strategy, for example, displaying different two-dimensional codes in a predetermined period.
  • the identification mark can be displayed in a variety of display modes, which can flexibly adapt to different application scenarios and can well meet different setting requirements.
  • the image acquisition unit is used for acquiring the target image according to the target object and the identification mark mapped on the target object.
  • the preset identification mark is displayed in the image display unit, and the identification mark is mapped on the target object as a light source.
  • the direction is located on the same side as the direction in which the identification mark is mapped to the target object by the display unit.
  • the image acquisition unit can photograph the target object with the identification mark mapped, and then collect the target image.
  • the image acquisition unit and the display unit are located on the same side of the target object, so that the setting of the identification mark and the acquisition of the target image can be completed in one mobile terminal, avoiding the use of redundant equipment, and the operation is simple and convenient.
  • the mobile terminal in order to make the authentication process more convenient and flexible, includes a mobile smart device, the display unit includes a screen, and the image acquisition unit includes a front camera.
  • Obtaining the target image through the image acquisition unit includes:
  • the mapping of the identification mark presented on the display unit on the target object through the display unit includes:
  • the identification mark displayed on the screen is mapped to the target object through the screen.
  • mobile smart devices may include mobile phones, tablets, or other mobile smart devices that can display images and perform image collection.
  • FIG. 3 is a schematic diagram of collecting an ID card image by using a mobile phone in an image recognition method proposed in an embodiment of this specification.
  • the mobile phone to collect the target image of the ID card as an example, where the display unit is the mobile phone screen, the image collection unit is the front camera of the mobile phone, and the target object is a copy of the ID card (ie the ID card displayed on the screen)
  • the identification mark is a QR code, as shown in Figure 3, the front camera of the mobile phone and the mobile phone screen are located on the same side of the ID card.
  • the two-dimensional code is mapped to the surface of the copy of the ID card, so that the two-dimensional code appears on the surface of the copy of the ID card (as shown in the dashed box in Figure 3).
  • use the front camera of the mobile phone to have a two-dimensional code on the surface Take a photo of the reproduced ID card of the person, and obtain a photo containing the reproduced ID card and a QR code as the target image.
  • the process of using the mobile phone to collect the ID card image is the same as the above process, and will not be repeated here.
  • the mapping of the identification mark and the collection of the target image can be completed by the mobile phone, which improves the applicability of the image recognition method proposed in the embodiment of this specification. Furthermore, by cooperating with the application in the mobile phone, it can meet the requirements of various application scenarios. Actual demand.
  • the method further includes: detecting the target object.
  • the target image of the target object includes:
  • the detection of the target object may include detecting whether the type of the target object meets the preset recognition requirements for recognition, for example, the recognition requirement is an ID card, and detecting whether the target object is an ID card; Detect whether the target object is in a predetermined position, for example, detect whether the edge of the target object exceeds the collection area.
  • the type of the target object when detecting whether the type of the target object meets the preset recognition requirements for recognition, it may include shooting the target object, acquiring a photo of the target object, and detecting whether there is a type of target certificate that meets the predetermined recognition requirement on the photo.
  • the type of the target object can be different types of documents such as an ID card and a driver's license, or a bank card.
  • the photo When detecting a photograph of a captured target object, the photo can be detected through a preset target object type detection model to determine whether the type of the target object in the photo meets the preset recognition requirements.
  • a preset target object type detection model For example, based on a neural network model, such as the SSD model (single shot multibox detector) in the convolutional neural network model, it uses multiple labeled photos containing preset target object types as input for training the SSD model. Perform training, and then use the trained SSD model to detect the captured photos of the target object to determine whether the type of the target object meets the preset recognition requirements.
  • the target object when detecting whether the target object is in a predetermined position, the target object can be photographed to obtain a photo of the target object, and whether the target object in the photo is in the predetermined position can be detected.
  • the neural network model such as the SSD model in the convolutional neural network model to identify the target object in the photo, and determine whether the edge of the target object is completely in the photo.
  • Multiple marked photos containing the target object at a predetermined position are used as input for training the SSD model.
  • the SSD model is trained, and then the trained SSD model is used to detect the photos of the target object to determine whether the target object is in The predetermined location.
  • the acquired target image can contain the complete target object, that is, the target image contains all parts of the target object, avoiding omissions and affecting subsequent recognition.
  • the detection of the target object can be carried out before the identification mark is mapped, that is, the identification mark is not mapped on the target object to be detected; it can also be carried out after the identification mark is mapped, that is, the identification mark is mapped. An identification mark is mapped on the target object.
  • the photo used for training the SSD model contains the target object but not the identification mark.
  • the photo used for training the SSD model contains both the target object and the identification mark.
  • the detection of the target object may include detecting the target object based on the identification mark when the identification mark is mapped on the target object.
  • detecting the target object based on the identification mark may include positioning by the identification mark, and detecting whether the target object is in the predetermined position based on the relative position of the identification mark and the target object.
  • a detection instruction is output according to the detection result.
  • it may include issuing a corresponding detection instruction as the prompt information according to the detection result, so as to understand the detection situation according to the detection instruction and perform subsequent operations.
  • a qualified instruction is issued to prompt the follow-up operation; when a qualified target certificate is not detected, a non-qualified instruction is issued to remind the target that an error has occurred and adjustment is required.
  • the prompt information can be displayed in many different ways. For example, different colors of lights can be used for prompting, different frequencies of vibration can be used for prompting, and different sounds can also be used for prompting.
  • the detection instructions are clearly and accurately output, and the detection instructions include voice broadcast and/or light instructions.
  • the detection instruction is a voice broadcast
  • the voice broadcast will prompt "the target certificate has been detected”; if the qualified target object is not detected, the detection will be repeated until the timeout expires.
  • the voice broadcast is used to cyclically prompt "the target certificate is not detected”; if the time is over, the voice broadcast is used to prompt "the target certificate is not detected, and exit over time".
  • the unqualified target object may include that the type of the target object does not meet a preset identification requirement, and may include that the target object is not in a predetermined position.
  • the detection instructions are output through voice prompts, so that the entire detection instructions can be clearly output, and the simplicity of the operation process is improved.
  • the detection indication is light
  • different colors of lights or light flashes of different frequencies can be used to prompt different detection conditions.
  • correspondingly acquiring the target image of the target object according to the detection result may include acquiring the target image according to the acquisition method corresponding to the detection result, for example, After the qualified target object is detected, the target image is automatically taken. It can also be taken according to the input shooting instruction after the qualified target object is detected. It can also be taken according to the input after the unqualified target object is detected. The shooting command takes the target image.
  • acquiring the target image of the target object includes: acquiring the target image within a preset acquisition time when the qualified target object is detected.
  • the target image when a qualified target object is detected, the target image should be acquired according to the qualified target object in a timely manner. For example, when the trained SSD model determines that the target object type conforms to When the recognition requirement is preset, the target image can be acquired within 1s after the determination, thereby avoiding the detection result being invalid due to an excessively long interval, and the acquired target image cannot be used, which affects subsequent recognition.
  • the collection time can be preset according to specific application scenarios, and there is no specific limitation here.
  • the target image is automatically collected, which can not only ensure the validity of the acquired target image, but also improve the recognition efficiency, and avoid the interference caused by redundant operations on the recognition process.
  • Step S203 Determine the attribute of the target object corresponding to the target image according to the preset mapping relationship according to the recognition feature.
  • the attributes of the target object may include the category to which the target object belongs.
  • the attribute of the target object when the target object is the original document, the attribute of the target object will be the original; when the target object is an electronic document forged based on the original document, the attribute of the target object will be a remake.
  • the preset mapping relationship includes the corresponding relationship between the recognition feature and the attribute of the target object, which is used to recognize the target image, and determine the attribute of the target object corresponding to the target image according to the recognition feature, that is, to determine whether the target image belongs to the original or the remake Pieces.
  • the preset mapping relationship can be obtained through a pre-trained remake recognition model, which is implemented based on a deep learning binary classification network.
  • the two categories include: remakes and originals; when training the remake recognition model, 500 images containing recognition features that have been labeled and classified are used as input for training.
  • the target image obtained in step S203 is input, and the recognition is performed through the photo-recognition model, and then the attributes of the target object are determined according to the recognition result.
  • the target image of the target object is acquired.
  • the target object is mapped with an identification mark.
  • the identification mark is used to form a corresponding identification feature in the target image, thereby using the identification mark on the target object.
  • the difference produced on the target image is formed on the target image with recognition features that can be used for recognition, and the recognition feature based on the difference is actively added to the recognition process, and then the target image corresponding to the target image is determined according to the recognition feature according to the preset mapping relationship
  • Attributes and mapping relationships include the corresponding relationship between the recognition features and the attributes of the target object, which effectively improves the accuracy of target image recognition.
  • Embodiment 2 for the same method as in Embodiment 1, the same symbols are used and the same description is omitted.
  • Fig. 4 is a flowchart of an authentication method proposed in an embodiment of the specification.
  • the authentication method of the embodiment of this specification includes the following steps:
  • Step S301 Acquire a first target image of the target object.
  • the target object is mapped with an identification mark, and the identification mark is used to form a corresponding identification feature in the first target image.
  • Step S303 Determine the attribute of the target object corresponding to the first target image according to the preset mapping relationship according to the recognition feature.
  • the mapping relationship includes the corresponding relationship between the recognition feature and the attribute of the target object.
  • Step S305 according to the determined result, authenticate the target object.
  • step S303 when it is determined in step S303 that the first target image is from the original, the target object is authenticated to confirm the authenticity of the target object.
  • the authentication of the target object may include extracting text information on the target object, comparing the text information with pre-stored information, and determining the authenticity of the target object.
  • the pre-stored information may include pre-filled information, and may include information pre-stored in a third-party system.
  • the method further includes:
  • the second target image After acquiring the first target image, acquire a second target image of the target object within a predetermined period of time, where the second target image is an image with no identification mark mapped on the target object;
  • the certification of target objects includes:
  • the authenticity of the target object is determined.
  • the target object in the second target image is a target object that is not mapped with an identification mark to avoid the influence of the identification mark on the authentication of the target object. Therefore, obtaining the second target image may be performed after step S303. At this time, first remove the identification mark on the target object, and then obtain the second target image; it can also be executed before step S303. For example, the second target image is obtained after the qualified target object is detected. The detection is based on the target object that is not mapped with the identification mark.
  • the second target image is acquired within a predetermined period of time. For example, when it is determined in step S303 that the first target image is from the original, it may be 0.5 seconds after the determination. Acquire the target image within.
  • Such a second target image acquisition method prevents the recognition result from being invalid due to a long interval, and ensures the accuracy of the second target image.
  • the second target image is acquired based on the recognition result of the first target image.
  • the target image is obtained according to the needs of authentication and is a resource that must be used in the subsequent authentication process, ensuring the effective use of resources.
  • the data processing strategy adopted may include using optical character recognition (OCR) technology to process the second target image containing the target object, thereby identifying and extracting information on the second target image, for example
  • OCR optical character recognition
  • the target object is an ID card
  • extract the name, birthday, and validity period of the ID on the ID card and then determine the authenticity of the target object based on the extracted information, for example, send the information extracted from the ID card to the public security system, Confirm the authenticity of the information, and then determine the authenticity of the ID card.
  • Fig. 5 is a flowchart of an authentication method proposed in an embodiment of this specification.
  • the authentication method of the embodiment of this specification includes the following steps:
  • step S401 the mobile phone outputs a detection start instruction, and then step S403 is executed.
  • the detection start instruction may include prompting the user to place the certificate to be authenticated in the collection area.
  • the output of the detection start instruction can be through voice broadcast; it can also be through the display of prompt information on the mobile phone screen.
  • step S403 it is judged whether the certificate is an ID card, if yes, step S405 is executed, and if otherwise, step S407 is executed.
  • the trained SSD model is used to detect the captured photo of the certificate to determine whether the certificate is an ID card.
  • Step S405 Display the two-dimensional code on the screen of the mobile phone, map the two-dimensional code to the ID card, and then perform step S4011.
  • Step S407 the mobile phone voice outputs a detection failure instruction, and it is judged whether the detection is overtime, if yes, step S409 is executed, and if otherwise, step S401 is executed.
  • Step S409 the mobile phone voice outputs a detection timeout instruction, and then enters the end.
  • Step S4011 the front camera of the mobile phone shoots the first target image, and then step S4013 is executed.
  • the first target image includes an ID card and an identification feature generated according to the two-dimensional code mapped on the ID card.
  • Step S4013 Determine the attribute of the target object corresponding to the first target image according to the preset mapping relationship according to the recognition feature. If the attribute of the target object corresponding to the first target image is the original ID card, then step S4015 is executed.
  • the first target image is recognized by using the trained remake recognition model to confirm whether the ID card is the original.
  • step S4015 the mobile phone screen stops displaying the two-dimensional code, the front camera captures the second target image, and then step S4017 is executed.
  • the second target image only includes the ID card.
  • Step S4017 the ID card is authenticated, and then the process ends.
  • the name, birthday, and validity period of the ID card are extracted through OCR technology, and the information is sent to the public security system to confirm the authenticity of the ID card. If the information is true, the authentication is passed. Authentication failed.
  • FIG. 6 is a schematic diagram of the structure of a picture recognition device proposed in an embodiment of this specification.
  • the picture recognition device described in the embodiment of this specification may include:
  • the acquiring unit 501 acquires a target image of the target object, where before acquiring the target image, the target object is mapped with an identification mark, and the identification mark is used to form a corresponding identification feature in the target image;
  • the recognition unit 503 determines the attribute of the target object corresponding to the target image according to the preset mapping relationship according to the recognition feature, and the mapping relationship includes the corresponding relationship between the recognition feature and the attribute of the target object.
  • the device before acquiring the target image, the device further includes: a detection unit that detects the target object,
  • the obtaining the target image of the target object includes:
  • the target image of the target object is correspondingly acquired according to the detection result.
  • the detecting the target object includes: when the identification mark is mapped on the target object, detecting the target object based on the identification mark.
  • the device further includes: an output unit that outputs a detection instruction according to the detection result.
  • the correspondingly acquiring the target image of the target object according to the detection result includes: when the qualified target object is detected, acquiring the target image within a preset acquisition time.
  • the device is used in a mobile terminal, and the mobile terminal includes a display unit and an image acquisition unit;
  • the step of acquiring the target image of the target object includes:
  • the step of mapping an identification mark on the target object includes:
  • the image acquisition unit and the display unit are located on the same side of the target object.
  • the above-mentioned at least one technical solution adopted in the embodiments of this specification can achieve the following beneficial effects: acquiring a target image of a target object, wherein before acquiring the target image, the target object is mapped with an identification mark, and the identification mark is used to form a corresponding image in the target image.
  • the identification feature that can be used for identification is formed on the target image, and the identification feature based on the difference is actively added to the identification process, and then according to the identification feature according to the preset
  • the mapping relationship determines the attributes of the target object corresponding to the target image, and the mapping relationship includes the corresponding relationship between the recognition feature and the attribute of the target object, which effectively improves the accuracy of target image recognition.
  • the embodiments of this specification also provide an electronic device, including at least one processor and a memory, the memory stores a program and is configured to execute the following steps by the at least one processor:
  • the attribute of the target object corresponding to the target image is determined according to the identification feature according to the preset mapping relationship, and the mapping relationship includes the corresponding relationship between the identification feature and the attribute of the target object.
  • embodiments of this specification also provide a computer-readable storage medium, including a program used in combination with an electronic device, and the program can be executed by a processor to complete the following steps:
  • the attribute of the target object corresponding to the target image is determined according to the identification feature according to the preset mapping relationship, and the mapping relationship includes the corresponding relationship between the identification feature and the attribute of the target object.
  • Fig. 7 is a schematic structural diagram of an authentication device proposed in an embodiment of this specification.
  • the authentication device described in the embodiment of this specification may include:
  • the acquiring unit 601 acquires a first target image of the target object, where before acquiring the target image, the target object is mapped with an identification mark, and the identification mark is used to form a corresponding identification feature in the first target image;
  • the recognition unit 603 determines the attribute of the target object corresponding to the first target image according to the preset mapping relationship according to the recognition feature, and the mapping relationship includes the corresponding relationship between the recognition feature and the attribute of the target object;
  • the authentication unit 605 authenticates the target object according to the determined result.
  • the device further includes:
  • the second target image After acquiring the first target image, acquire a second target image of the target object within a predetermined period of time, where the second target image is an image to which the identification mark is not mapped on the target object;
  • the authentication of the target object includes:
  • the authenticity of the target object is determined.
  • the embodiments of this specification also provide an electronic device, including at least one processor and a memory, the memory stores a program and is configured to execute the following steps by the at least one processor:
  • the mapping relationship includes the corresponding relationship between the identification feature and the attribute of the target object
  • the target object is authenticated.
  • embodiments of this specification also provide a computer-readable storage medium, including a program used in combination with an electronic device, and the program can be executed by a processor to complete the following steps:
  • the mapping relationship includes the corresponding relationship between the identification feature and the attribute of the target object
  • the target object is authenticated.
  • a programmable logic device Programmable Logic Device, PLD
  • FPGA Field Programmable Gate Array
  • HDL Hardware Description Language
  • ABEL Advanced Boolean Expression Language
  • AHDL Altera Hardware Description Language
  • HDCal JHDL
  • Lava Lava
  • Lola MyHDL
  • PALASM RHDL
  • Verilog Verilog
  • the controller can be implemented in any suitable manner.
  • the controller can take the form of, for example, a microprocessor or a processor and a computer-readable medium storing computer-readable program codes (such as software or firmware) executable by the (micro)processor. , Logic gates, switches, application specific integrated circuits (ASICs), programmable logic controllers and embedded microcontrollers.
  • controllers include but are not limited to the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20 and Silicon Labs C8051F320, the memory controller can also be implemented as a part of the memory control logic.
  • controller in addition to implementing the controller in a purely computer-readable program code manner, it is entirely possible to program the method steps to make the controller use logic gates, switches, application specific integrated circuits, programmable logic controllers and embedded The same function can be realized in the form of a microcontroller, etc. Therefore, such a controller can be regarded as a hardware component, and the devices included in it for implementing various functions can also be regarded as a structure within the hardware component. Or even, the device for realizing various functions can be regarded as both a software module for realizing the method and a structure within a hardware component.
  • a typical implementation device is a computer.
  • the computer may be, for example, a personal computer, a laptop computer, a cell phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or Any combination of these devices.
  • the embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.
  • a computer-usable storage media including but not limited to disk storage, CD-ROM, optical storage, etc.
  • These computer program instructions can also be stored in a computer-readable memory that can guide a computer or other programmable data processing equipment to work in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture including the instruction device.
  • the device implements the functions specified in one process or multiple processes in the flowchart and/or one block or multiple blocks in the block diagram.
  • These computer program instructions can also be loaded on a computer or other programmable data processing equipment, so that a series of operation steps are executed on the computer or other programmable equipment to produce computer-implemented processing, so as to execute on the computer or other programmable equipment.
  • the instructions provide steps for implementing functions specified in a flow or multiple flows in the flowchart and/or a block or multiple blocks in the block diagram.
  • the computing device includes one or more processors (CPU), input/output interfaces, network interfaces, and memory.
  • processors CPU
  • input/output interfaces network interfaces
  • memory volatile and non-volatile memory
  • the memory may include non-permanent memory in computer readable media, random access memory (RAM) and/or non-volatile memory, such as read-only memory (ROM) or flash memory (flash RAM). Memory is an example of computer readable media.
  • RAM random access memory
  • ROM read-only memory
  • flash RAM flash memory
  • Computer-readable media include permanent and non-permanent, removable and non-removable media, and information storage can be realized by any method or technology.
  • the information can be computer-readable instructions, data structures, program modules, or other data.
  • Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, CD-ROM, digital versatile disc (DVD) or other optical storage, Magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media can be used to store information that can be accessed by computing devices. According to the definition in this article, computer-readable media does not include transitory media, such as modulated data signals and carrier waves.
  • program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types.
  • This application can also be practiced in distributed computing environments. In these distributed computing environments, remote processing devices connected through a communication network perform tasks.
  • program modules can be located in local and remote computer storage media including storage devices.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Multimedia (AREA)
  • Image Analysis (AREA)

Abstract

本说明书实施例公开了一种图像识别方法、装置、设备和认证方法、装置、设备，其方案是，获取目标对象的目标图像，其中，在获取目标图像前，目标对象上映射有识别标识，识别标识用于在目标图像中形成对应的识别特征；根据识别特征按预设的映射关系确定目标图像对应的目标对象的属性，映射关系包括识别特征与目标对象的属性的对应关系。

Description

图像识别方法、装置、设备和认证方法、装置、设备 技术领域
本说明书实施例涉及计算机技术领域，尤其涉及一种图像识别方法、装置、设备和认证方法、装置、设备。
背景技术
随着技术的进步，越来越多的业务可以通过远程办理，比如账户开户(如金融账户)，在办理远程业务时，需要对待认证对象的真伪进行认证，比如要求用户手持身份证原件摄像进行认证，又比如要求与用户视频面谈进行认证，这样的认证方式因需要人工参与，已很难满足业务量大的应用场景，因而已逐步出现对待认证对象的电子物进行验证，比如要求用户拍摄身份证照片并在拍摄后上传该身份证照片，但在远程自动认证中，往往对翻拍的照片(即翻拍件)很难识别，或者识别的准确性不高，这样就很容易被不法分子通过翻拍伪造的电子证件来办理这些远程业务，其中，伪造的电子证件是指将物理证件，通过拍照、扫描手段电子化后，在得到的电子版证件图像上进行信息篡改的行为，如使用PS(Photoshop)软件篡改，而伪造的电子证件由于具有伪造目标易获取、造假成本低、造假工具易获取且效果逼真等特点，使得伪造电子证件经常出现在远程认证证件的过程中，对远程业务的安全带来很大的威胁。
发明内容
有鉴于此，本说明书实施例提供了一种图像识别方法、装置、设备和认证方法、装置、设备，用于图像的类别进行识别。
本说明书实施例采用下述技术方案：
本说明书实施例提供一种图像识别方法，包括：
获取目标对象的目标图像，其中，在获取目标图像前，目标对象上映射有识别标识，识别标识用于在目标图像中形成对应的识别特征；
根据识别特征按预设的映射关系确定目标图像对应的目标对象的属性，映射关系包括识别特征与目标对象的属性的对应关系。
本说明书实施例提供一种认证方法，包括：
获取目标对象的第一目标图像，其中，在获取第一目标图像前，目标对象上映射有识别标识，识别标识用于在第一目标图像中形成对应的识别特征；
根据识别特征按预设的映射关系确定第一目标图像对应的目标对象的属性，映射关系包括识别特征与目标对象的属性的对应关系；
根据确定的结果，对目标对象进行认证。
本说明书实施例提供一种图像识别装置，包括：
获取单元，获取目标对象的目标图像，其中，在获取目标图像前，目标对象上映射有识别标识，识别标识用于在目标图像中形成对应的识别特征；
识别单元，根据识别特征按预设的映射关系确定目标图像对应的目标对象的属性，映射关系包括识别特征与目标对象的属性的对应关系。
本说明书实施例提供一种认证装置，包括：
获取单元，获取目标对象的第一目标图像，其中，在获取第一目标图像前，目标对象上映射有识别标识，识别标识用于在第一目标图像中形成对应的识别特征；
识别单元，根据识别特征按预设的映射关系确定第一目标图像对应的目标对象的属性，映射关系包括识别特征与目标对象的属性的对应关系；
认证单元，根据确定的结果，对目标对象进行认证。
本说明书实施例提供一种电子设备，包括：至少一个处理器及存储器，所述存储器存储有程序，并且被配置成由至少一个处理器执行以下步骤：
获取目标对象的目标图像，其中，在获取目标图像前，目标对象上映射有识别标识，识别标识用于在目标图像中形成对应的识别特征；
根据识别特征按预设的映射关系确定目标图像对应的目标对象的属性，映射关系包括识别特征与目标对象的属性的对应关系。
本说明书实施例提供一种电子设备，包括：至少一个处理器及存储器，所述存储器存储有程序，并且被配置成由至少一个处理器执行以下步骤：
获取目标对象的第一目标图像，其中，在获取第一目标图像前，目标对象上映射有识别标识，识别标识用于在第一目标图像中形成对应的识别特征；
根据识别特征按预设的映射关系确定第一目标图像对应的目标对象的属性，映射关 系包括识别特征与目标对象的属性的对应关系；
根据确定的结果，对目标对象进行认证。
本说明书实施例采用的上述至少一个技术方案能够达到以下有益效果：
获取目标对象的目标图像，其中，在获取目标图像前，目标对象上映射有识别标识，识别标识用于在目标图像中形成对应的识别特征，从而利用识别标识在目标对象上产生的差异，在目标图像上形成能够用于识别的识别特征，主动地为识别过程增加基于差异性产生的识别特征，再根据识别特征按预设的映射关系确定目标图像对应的目标对象的属性，映射关系包括识别特征与目标对象的属性的对应关系，有效地提高了目标图像识别的准确性。
附图说明
此处所说明的附图用来提供对本说明书实施例的进一步理解，构成本说明书实施例的一部分，本说明书的示意性实施例及其说明用于解释本申请，并不构成对本申请的不当限定。在附图中：
图1为本说明书实施例提出的一种图像识别方法的应用示意图。
图2为本说明书实施例提出的一种图像识别方法的流程图。
图3为本说明书实施例提出一种图像识别方法中采用手机采集身份证图像的采集示意图。
图4为本说明书实施例提出的一种认证方法的流程图。
图5为本说明书实施例提出的一种认证方法的流程图。
图6为本说明书实施例提出的一种图像识别装置的结构示意图。
图7为本说明书实施例提出的一种认证装置的结构示意图。
具体实施方式
为使本申请的目的、技术方案和优点更加清楚，下面将结合本说明书具体实施例及相应的附图对本申请技术方案进行清楚、完整地描述。显然，所描述的实施例仅是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。
现阶段，对待认证的电子物，例如，根据待认证的证件获取的待识别图片，的识别可采用以下方法进行识别：根据待认证对象获取待识别图片，从图片中提取识别特征，根据这些识别特征对图片进行分类识别，进而确定待识别的对象是否来源于原件。
上述方法中仅通过对根据待认证对象获取的待识别图片进行识别处理，能够从待识别图片获取的识别信息是有限的，导致识别效果不佳，影响判断的准确性。
发明人发现，对待认证对象进行认证的时，需要对待认证对象进行拍照，获取待认证的电子物，如待识别图片，此时，若待认证的对象为根据待认证对象伪造的电子件，例如伪造的电子证件，在伪造完成后必须将伪造的电子件借助某种物理方式呈现出来，例如，通过屏幕显示或打印显示，才能通过二次翻拍的方式获取待识别图片，进而将其输入认证系统进行认证，因此，本说明书实施例提出的图像识别方法、装置、设备和认证方法、装置、设备，充分利用了伪造的电子件需要通过屏幕显示或打印显示才能用于认证的特点，通过对这类根据伪造的电子件进行二次翻拍获得的待识别图像进行识别，有效地防控在认证过程中使用伪造的电子证件。
本说明书实施例中的图像识别过程的应用示意图可如图1所示，包括待识别的目标对象10，例如身份证，映射在目标对象上的识别标识20，例如二维码，以及根据目标对象10与识别标识20获得的目标图像30。
进行识别时，根据映射有识别标识20的目标对象10获取目标图像30，该目标图像30上具有根据识别标识20形成的识别特征，获取到目标图像30后，根据识别特征按预设的映射关系对目标图像30进行识别，进而确定目标图像30对应的目标对象的属性。
具体实施中，识别标识20的映射以及目标图像30的采集、识别和确定可以是由同一主体执行，也可以是由不同的主体执行。
本说明书实施例提供的图像识别方法由同一主体执行时，例如由终端执行，过程如下：
通过终端在识别对象10上映射识别标识20，获取目标图像30，对目标图像30进行识别，确定目标图像30对应的目标对象的属性。
本说明书实施例提供的图像识别方法由不同主体执行时，例如由终端和服务端共同执行，过程如下：
通过终端在识别对象10上映射识别标识20，获取目标图像30，将目标图像30发送至服务端进行识别，再将识别结果发送至终端以确定目标图像30对应的目标对象的属 性。
上述共同执行的过程还可以是，通过第一终端在识别对象10上映射识别标识20，通过第二终端获取目标图像30，将目标图像30发送至服务端进行识别，再将识别结果发送至第二终端以确定目标图像30对应的目标对象的属性。
其中，终端包括能够运行软件的终端设备，包括但不限于计算机、平板电脑、手机、银行的客户端设备等。
本说明书实施例提出的图像识别方法，其整体思路是，利用伪造的电子件必须通过屏幕显示或打印显示呈现出来，使得呈现出的伪造的电子件与原件具有不同的表面材质的特性，通过在目标对象上映射识别标识，基于原件与呈现出的伪造得电子件表面材质不同的原因，该识别标识显示在目标对象上时会产生差异，进而在获取带有识别标识的目标对象对应的目标图像时，识别标识能够在目标图像上形成对应的识别特征，例如识别特征可以包括根据识别标识在目标对象上产生的差异形成的特征，从而主动地为识别过程增加识别特征，最后基于这种利用差异性主动增加的识别特征，通过已训练出来的映射关系，确定出目标图像对应的目标对象的属性，例如，目标图像可以是根据原件获得的图像，也可以是根据显示出的伪造的电子件获得图像，有效地提高了目标图像的识别准确性。
本说明书实施例提出的图像的识别方法，获取目标对象的目标图像，其中，在获取目标图像前，目标对象上映射有识别标识，识别标识用于在目标图像中形成对应的识别特征，充分利用了识别标识在不同种类的目标对象上产生的差异，基于这种差异在目标图像上形成可被识别出的识别特征，为后续识别提供丰富的识别信息，同时通过映射的方式设置识别标识，避免了对目标对象的损伤；然后，根据识别特征按预设的映射关系识别目标图像的目标对象的属性，基于识别特征进行识别判断，有效地提高了识别的准确性。这样的识别过程，通过在目标对象上映射的识别标识的方式，主动为识别过程增加易于识别的特征，并基于新增的识别特征对图片进行识别，既提高了准确性又提高了识别效率，避免无关特征对识别过程和识别结果的干扰。
上述应用场景仅是为了便于理解本申请而示出，本说明书的实施方式在此方面不受任何限制。相反，本说明书的实施方式可以应用于适用的任何场景。
以下，参照附图对本说明书的图像识别方法、装置、设备和认证方法、装置、设备进行详细阐述。
实施例1
图2为本说明书实施例提出的一种图像识别方法的流程图。
如图2所示，本说明书实施例中图像识别方法包括以下步骤：
步骤S201，获取目标对象的目标图像。
其中，在获取目标图像前，目标对象上映射有识别标识，识别标识用于在目标图像中形成对应的识别特征。
本说明书实施例中，目标对象包括用于生成目标图像的对象。
例如，在需要识别身份的应用场景中，目标对象可以是用于表征身份的证件，比如身份证、驾驶证、护照等，此时，证件可能为证件的原件，也可能为根据原件而伪造的电子件(如，证件的翻拍件)。
又如，在需要识别是否真实持有银行卡的应用场景中，目标对象也可以是各种形式的银行卡，此时，银行卡可能为银行卡的原件，也可能为根据原件而伪造的电子件(如，银行卡的翻拍件)。
需要说明的是，翻拍件为通过物理方式再次呈现出来的翻拍件，例如，屏幕显示的翻拍件或打印显示的翻拍件。
识别标识包括映射在目标对象上，用于产生供识别的特征的标识，例如，识别标识可以是特定的图案等。
在实际操作的过程中，识别标识被映射在目标对象表面，可以是通过带有识别标识的光源向外发射光线，目标对象靠近识别标识时，光线在目标对象的表面发生反射，进而在目标对象表面映射出可以形成识别特征的识别标识。
这样的识别标识映射方法，既能够根据实际需要的调整识别标识，满足复杂的实际使用需求，又能够避免对目标对象造成破坏，同时还能够简单、快捷、清晰地将识别标识设置于目标对象上。
识别特征包括在目标图像中形成的与识别标识对应的，能够被识别的特征。
具体地，根据目标对象不同，用于获取目标图像的目标对象具有不同的表面材质，因此，将识别标识映射在目标对象上时，由于表面材质的不同，识别标识在不同的表面上会产生差异。例如，当目标对象为身份证原件时，表面材质为防伪镀膜，当目标对象为屏幕显示的身份证翻拍件时，表面材质为液晶显示屏，此时，识别标识在两种不同材 质的表面会产生明显的差异，获取到的目标图像中的与识别标识对应的识别特征同样会保留这些差异，进而识别特征能够基于这些差异而被用于识别。
在一种应用示例中，为了能够更准确地进行识别，识别标识包括包含定位特征的图案，该定位特征用于包含定位特征的图案的定位，具体地，通过图案中包含的定位特征能够快速地对图案进行定位，即能够根据定位特征快速地对识别标识进行定位，进而进行识别。
例如，上述包含定位特征的图案可以是圆形图案、中心对称的多边形图案等。
采用了包含定位特征的图案作为识别标识，使得后续的识别过程中能够基于定位特征对识别标识进行定位，确定识别特征的位置，进而保证识别特征能够更为准确、快捷的被用于识别。
本说明书实施例中，获取到的目标图像中包括目标对象和与识别标识对应的识别特征。
在实际应用过程中，为了能够简捷快速地获取目标图像，获取目标图像的方式可以包括拍摄获取，例如，采用摄像头对目标对象和识别标识进行拍摄，例如，拍摄照片或视频，再获取拍摄得到的一张照片或视频中的一张图片作为目标图像；也可以包括扫描获取，例如，采用扫描仪对目标对象和识别标识进行扫描，获取扫描到的图像作为目标图像。
进一步地，获取到的目标图像中的识别特征包括颜色特征、纹理特征、形状特征和空间关系特征中的任意一种或多种。
例如，当目标对象为身份证时，识别特征在身份证原件和屏幕显示或打印显示的身份证翻拍件上显示时，由于原件与屏幕或纸质打印件的表面材质的差异性，使得映射在其上的识别标识会产生灰度、色彩上的差异，纹理上的差异，轮廓上的形变差异以及与身份证相对位置的差异等，基于这些差异使得识别标识能够在目标图像上能够形成对应的识别特征。
在一种应用示例中，为了使整个识别过程能够便于操作，避免使用复杂的设备，本说明书实施例中的图像识别方法可以用于移动终端中，移动终端包括显示单元和图像采集单元；
获取目标对象的目标图像的步骤包括：
通过图像采集单元获取目标图像；
在目标对象上映射识别标识的步骤包括：
通过显示单元将在显示单元上呈现的识别标识映射在目标对象上；
其中，图像采集单元与显示单元位于目标对象的同一侧。
本说明书实施例中，移动终端可以包括手机、平板、计算机、银行客户端或其他能够显示图像并进行图像采集的移动设备。
显示单元用于呈现识别标识，将识别标识映射到目标对象上。
识别标识可以包括随机产生的识别标识，例如圆形图标、方形图标等；还可以是包含预置有特定信息的识别标识，例如商标、缩写、二维码、条形码、编号等。
显示单元对识别标识的呈现可以包括显示相同的识别标识，例如固定显示银行的缩写作为识别标识；也可以是根据预定的显示策略显示识别标识，例如按预定的周期显示不同的二维码。
识别标识可以采用多种显示方式进行显示，既能够灵活地适应不同的应用场景，又能够很好地满足不同的设置需求。
图像采集单元用于根据目标对象和目标对象上映射的识别标识采集目标图像。
具体地，在移动终端进行目标图像的采集过程中，图像显示单元中显示出预设的识别标识，并作为光源将识别标识映射在目标对象上，此时，由于图像采集单元采集目标图像的采集方向，与显示单元将识别标识映射到目标对象上的方向位于同一侧，图像采集单元能够将映射有识别标识的目标对象拍摄下来，进而采集到目标图像。
图像采集单元与显示单元位于目标对象的同一侧，使得能够识别标识的设置与目标图像的采集能够在一台移动终端中完成，避免了多余设备的使用，操作简单方便。
在一种应用示例中，为了能够使得认证的过程更为便捷、灵活，移动终端包括移动智能设备，显示单元包括屏幕，图像采集单元包括前置摄像头，
通过图像采集单元获取目标图像包括：
通过前置摄像头获取目标图像；
通过显示单元将在显示单元上呈现的识别标识映射在目标对象上包括：
通过屏幕将在屏幕上显示的识别标识映射在目标对象上。
本说明书实施例中，移动智能设备可以包括手机、平板或其他能够显示图像并进行图像采集的移动智能设备。
图3为本说明书实施例提出的一种图像识别方法中采用手机采集身份证图像的采集示意图。
具体地，以手机对身份证进行目标图像的采集为例进行说明，其中，显示单元为手机屏幕，图像采集单元为手机前置摄像头，目标对象为身份证翻拍件(即屏幕显示的身份证)，识别标识为二维码，如图3所示，手机前置摄像头与手机屏幕位于身份证同一侧，这时可手持手机对着屏幕显示的身份证进行拍摄，通过手机屏幕显示的二维码，把二维码映射到身份证翻拍件表面，使得身份证翻拍件表面出现二维码(如图3中虚线框内所示)，此时，利用手机前置摄像头，对表面具有二维码的身份证翻拍件进行拍照，获取包含身份证翻拍件和二维码的照片作为目标图像。
需要说明的是，当目标对象为身份证原件时，采用手机采集身份证图像的过程与上述过程相同，此处不再一一赘述。
通过手机即可完成识别标识的映射与目标图像的采集，提高了本说明书实施例提出的图像识别方法的适用性，进一步地，通过与手机中的应用的相互配合，即可满足各个应用场景的实际需求。
在一种应用示例中，为了保证能够准确地获取到有效的目标图像，在获取目标图像前，所述的方法还包括：检测目标对象。
获取目标对象的目标图像包括：
根据检测结果对应地获取目标对象的目标图像。
具体地，在获取目标图像之前，对目标对象的检测可以包括对目标对象的类型是否符合识别预设的识别要求进行检测，例如识别要求为身份证，检测目标对象是否为身份证；也可以包括对目标对象是否处于预定的位置进行检测，例如检测目标对象的边缘是否超出采集区域。
例如，当对目标对象的类型是否符合识别预设的识别要求进行检测时，可以包括对目标对象进行拍摄，获取目标对象的照片，检测照片上是否存在符合预定识别需求的目标证件的类型。其中，目标对象的类型可以是身份证、驾驶证等不同类型的证件，也可以是银行卡。
对拍摄到的目标对象的照片进行检测时，可以通过预设的目标对象类型检测模型对照片进行检测，确定照片中目标对象的类型是否符合预设的识别需求。例如，基于神经网络模型，如卷积神经网络模型中的SSD模型(single shot multibox detector)进行识别，利用多个标记过的包含预设目标对象类型的照片作为训练SSD模型的输入，对SSD模型进行训练，再利用训练好的SSD模型对拍摄到的目标对象的照片进行检测，确定目标对象的类型是否符合预设的识别要求。
通过提前对目标对象的类型进行检测，避免了后续识别过程中由于使用了错误的目标对象造成的识别失败，提高了识别效率。
又如，当对目标对象是否处于预定的位置进行检测时，可以通过对目标对象进行拍摄，获取目标对象的照片，并检测照片中的目标对象是否处于预定的位置。
对拍摄到的目标对象的照片进行检测时，同样可以使用神经网络模型，如卷积神经网络模型中的SSD模型对照片中的目标对象进行识别，确定目标对象的边缘是否完全处于照片中，利用多个标记过的包含处于预定位置的目标对象的照片作为训练SSD模型的输入，对SSD模型进行训练，再利用训练好的SSD模型对拍摄到的目标对象的照片进行检测，确定目标对象是否处于预定的位置。
由于对目标对象是否处于预定的位置进行检测，使得获取到的目标图像能够包含完整的目标对象，即目标图像中包含目标对象的所有部分，避免出现遗漏，影响后续识别。
需要说明的是，对目标对象的检测可以是在映射有识别标识之前进行的，即待检测的目标对象上未映射有识别标识；也可以时在映射有识别标识之后进行的，即待检测的目标对象上映射有识别标识。
当在映射识别标识之前检测目标对象时，对SSD模型进行训练采用的照片中包含目标对象，不包含识别标识。
当在映射识别标识之后检测目标对象时，对SSD模型进行训练采用的照片中同时包含目标对象和识别标识。
在一种应用示例中，当在映射识别标识之后检测目标对象时，即对目标对象的检测可以包括在当目标对象上映射有识别标识时，基于识别标识检测目标对象。
例如，当对目标对象是否处于预定的位置进行检测时，基于识别标识检测目标对象可以包括通过识别标识进行定位，基于识别标识与目标对象的相对位置检测目标对 象是否处于预定的位置。
在一种应用示例中，为了能够便于确定检测的结果，提高检测效率，在检测目标对象中，根据检测结果输出检测指示。
具体地，可以包括根据检测的结果发出相应的检测指示作为提示信息，以便于根据检测指示了解检测情况，执行后续操作。
例如，当检测到合格目标证件时，发出合格指示，以提示进入后续操作；当未检测到合格的目标证件时，发出不合格指示，以提示目标对象发生错误，需要进行调整。
需要说明的是，提示信息可以通过多种不同的方式展现，例如，可以采用不同颜色的灯光进行提示，也可以采用不同频率的震动进行提示，还可以采用不同的声音进行提示。
在一种应用示例中，为了适应实际使用中的复杂的检测情况，清晰准确地输出检测指示，检测指示包括语音播报和/或灯光指示。
例如，当检测指示为语音播报时，若检测到了合格的目标对象，则用语音播报的形式提示“已检测到目标证件”；若未检测到合格的目标对象，则循环检测直到超时，并在在循环检测的过程中通过语音播报的形式循环提示“未检测到目标证件”；如果超时，则用语音播报的形式提示“未能检测到目标证件，超时退出”。其中，未检测到合格的目标对象可以包括目标对象的种类不符合预设的识别需求，可以包括目标对象未处于预定的位置。
通过语音提示的方式输出检测指示，使得整个检测指令能够的得到清晰的输出，提高了操作过程的简便性。
当检测指示为灯光时，可以通过不同颜色的灯光或不同频率的灯光闪烁来提示不同的检测情况。
进一步地，预设有多种的目标图像获取方式，在获得检测结果后，根据检测结果对应地获取目标对象的目标图像可以包括根据与检测结果对应的获取方式获取目标图像，例如，可以是在检测到合格的目标对象后，自动拍摄目标图像，也可以是在检测到合格的目标对象后，根据输入的拍摄指令拍摄目标图像，还可以是在检测到不合格的目标对象后，根据输入的拍摄指令拍摄目标图像。
在一种应用示例中，为了保证目标图像是根据检测合格的目标对象获取到的， 获取目标对象的目标图像包括：检测到合格的目标对象时，在预设的采集时间内获取目标图像。
具体地，为了保证获取到的目标图像的准确性，在检测到合格的目标对象时，应当及时地根据检测合格的目标对象获取目标图像，例如，当训练好的SSD模型确定目标对象的种类符合预设的识别要求时，可以是在确定后的1s内获取目标图像，进而避免了间隔过长导致检测结果失效，获取到的目标图像无法使用，影响后续识别。
需要说明的是，本说明书实施例中，采集时间可根据具体应用场景需要进行预设，这里不做具体限定。
基于对目标对象的检测结果，自动进行目标图像的采集，既能够保证获取到的目标图像的有效性，又提高了识别效率，避免了多余操作对识别过程造成的干扰。
步骤S203，根据识别特征按预设的映射关系确定目标图像对应的目标对象的属性。
其中，目标对象的属性可以包括目标对象所归属的种类。
以前述目标对象身份证件为例，当目标对象为证件的原件时，目标对象的属性将为原件；当目标对象为根据证件的原件而伪造的电子件时，目标对象的属性将为翻拍件。
其中，预设的映射关系包括识别特征与目标对象的属性的对应关系，用于对目标图像进行识别，根据识别特征确定目标图像对应的目标对象的属性，即确定目标图像属于原件，还是属于翻拍件。预设的映射关系可以通过预先训练好的翻拍识别模型获得，该翻拍识别模型基于深度学习的二分类网络来实现。具体地，两个分类包括：翻拍件和原件；对翻拍识别模型进行训练时，采用分别标注过分类的包含识别特征的图像各500张作为输入进行训练。识别时，输入步骤S203中获得的目标图像，通过翻拍识别模型进行识别，进而根据识别结果确定目标对象的属性。
本说明书实施例中，获取目标对象的目标图像，其中，在获取目标图像前，目标对象上映射有识别标识，识别标识用于在目标图像中形成对应的识别特征，从而利用识别标识在目标对象上产生的差异，在目标图像上形成能够用于识别的识别特征，主动地为识别过程增加基于差异性产生的识别特征，再根据识别特征按预设的映射关系确定目标图像对应的目标对象的属性，映射关系包括识别特征与目标对象的属性的对应关系，有效地提高了目标图像识别的准确性。
实施例2
在实施例2中，对于与实施例1中相同的方法，使用相同的符号并省略相同的说明。
图4为本说明书实施例提出的一种认证方法的流程图。
如图4所示，本说明书实施例的认证方法包括以下步骤：
步骤S301，获取目标对象的第一目标图像。
其中，在获取第一目标图像前，目标对象上映射有识别标识，识别标识用于在第一目标图像中形成对应的识别特征。
步骤S303，根据识别特征按预设的映射关系确定第一目标图像对应的目标对象的属性。
其中，映射关系包括识别特征与目标对象的属性的对应关系。
步骤S305，根据确定的结果，对目标对象进行认证。
具体地，当步骤S303中确定了第一目标图像来自于原件时，对目标对象进行认证，以确认目标对象的真伪。
例如，对目标对象的认证可以包括提取目标对象上的文字信息，将文字信息与预存的信息进行比较，判断目标对象的真伪。其中，预存的信息可以包括预先填写的信息，可以包括通过第三方系统中预存的信息。
在一种应用示例中，为了保证认证过程的准确性，对所述的方法还包括：
在获取第一目标图像后，在预定时间段内获取目标对象的第二目标图像，第二目标图像为目标对象上未映射识别标识的图像；
对目标对象进行认证包括：
按预设的数据处理策略，对第二目标图像进行数据处理；
根据数据处理的结果，确定目标对象的真伪。
本说明书实施例中，在第二目标图像中目标对象为未映射有识别标识的目标对象，以避免识别标识对于目标对象认证的影响，因此，获取第二目标图像可以是在步骤S303之后执行，此时，先将目标对象上的识别标识去除，再获取第二目标图像；也可以是在步骤S303之前执行，例如，检测到合格的目标对象后获取第二目标图像，此时， 对目标对象的检测是基于未映射有识别标识的目标对象进行的。
例如，可以是在确定了第一目标图像来自于原件后，在预定时间段获取第二目标图像，如当步骤S303中确定了第一目标图像来自于原件时，可以是在确定后的0.5s内获取目标图像。
这样的第二目标图像获取方法，避免了间隔过长导致识别结果失效，保证了第二目标图像的准确性，同时，第二目标图像的获取是基于第一目标图像的识别结果进行的，第二目标图像是根据认证需要而获取的，是后续认证过程中必须使用的资源，确保了资源的有效利用。
进一步地，对目标对象进行认证时，采用的数据处理策略可以包括使用光学字符识别(OCR)技术对包含目标对象的第二目标图像进行处理，从而识别并提取第二目标图像上的信息，例如，目标对象为身份证时，提取身份证上的姓名、生日、证件有效期等信息，再根据提取到的信息确定目标对象的真伪，例如，将身份证上提取到的信息发送至公安系统，确认信息的真伪，进而判断身份证的真伪。
以下以手机中的认证应用对身份证的真伪进行认证的过程为例，对本说明书提供的认证方法进行说明：
图5为本说明书实施例提出的一种认证方法的流程图。
如图5所示，本说明书实施例的认证方法包括以下步骤：
步骤S401，手机输出检测开始指令，然后执行步骤S403。
其中，检测开始指令可以包括提示用户将待认证的证件置于采集区域。
检测开始指令的输出可以是通过语音播报的方式；也可以是通过手机屏幕上的显示提示信息的方式。
步骤S403，判断证件是否为身份证，若是则执行步骤S405，若否则执行步骤S407。
本说明书实施例中，通过对证件进行拍照，利用训练好的SSD模型对拍摄到的证件照片进行检测，判断证件是否为身份证。
步骤S405，在手机屏幕上显示二维码，将二维码映射到身份证上，然后执行步骤S4011。
步骤S407，手机语音输出检测失败指令，判断检测是否超时，若是则执行步骤 S409，若否则执行步骤S401。
步骤S409，手机语音输出检测超时指令，然后进入结束。
步骤S4011，手机前置摄像头拍摄第一目标图像，然后执行步骤S4013。
其中，第一目标图像包括身份证和根据映射在身份证上的二维码生成的识别特征。
步骤S4013，根据识别特征按预设的映射关系确定第一目标图像对应的目标对象的属性，若第一目标图像对应的目标对象的属性为身份证原件，则执行步骤S4015。
本说明书实施例中，利用训练好的翻拍识别模型对第一目标图像进行识别，确认身份证是否为原件。
步骤S4015，手机屏幕停止显示二维码，前置摄像头拍摄第二目标图像，然后执行步骤S4017。
其中，第二目标图像中仅包括身份证。
步骤S4017，对身份证进行认证，然后进入结束。
本说明书实施例中，通过OCR技术提取身份证上的姓名、生日、证件有效期等信息，将这些信息发送至公安系统，确认身份证的真伪，若信息真实，则认证通过，若信息虚假则认证失败。
实施例3
图6为本说明书实施例提出的一种图片识别装置的结构示意图。
基于同一个申请构思，本说明书实施例记载的图片识别装置可以包括：
获取单元501，获取目标对象的目标图像，其中，在获取目标图像前，目标对象上映射有识别标识，识别标识用于在目标图像中形成对应的识别特征；
识别单元503，根据识别特征按预设的映射关系确定目标图像对应的目标对象的属性，映射关系包括识别特征与目标对象的属性的对应关系。
可选地，在获取所述目标图像前，所述装置还包括：检测单元，检测所述目标对象，
所述获取目标对象的目标图像包括：
根据检测结果对应地获取所述目标对象的目标图像。
可选地，所述检测所述目标对象包括：当所述目标对象上映射有所述识别标识时，基于所述识别标识检测所述目标对象。
可选地，在所述检测所述目标对象中，所述装置还包括：输出单元，根据检测结果输出检测指示。
可选地，所述根据检测结果对应地获取所述目标对象的目标图像包括：检测到合格的所述目标对象时，在预设的采集时间内获取所述目标图像。
可选地，所述装置用于移动终端中，所述移动终端包括显示单元和图像采集单元；
所述获取所述目标对象的目标图像的步骤包括：
通过所述图像采集单元获取所述目标图像；
在所述目标对象上映射识别标识的步骤包括：
通过所述显示单元将在所述显示单元上呈现的所述识别标识映射在所述目标对象上；
其中，所述图像采集单元与所述显示单元位于所述目标对象的同一侧。
本说明书实施例采用的上述至少一个技术方案能够达到以下有益效果：获取目标对象的目标图像，其中，在获取目标图像前，目标对象上映射有识别标识，识别标识用于在目标图像中形成对应的识别特征，从而利用识别标识在目标对象上产生的差异，在目标图像上形成能够用于识别的识别特征，主动地为识别过程增加基于差异性产生的识别特征，再根据识别特征按预设的映射关系确定目标图像对应的目标对象的属性，映射关系包括识别特征与目标对象的属性的对应关系，有效地提高了目标图像识别的准确性。
基于同一发明构思，本说明书实施例还提供一种电子设备，包括至少一个处理器及存储器，所述存储器存储有程序，并且被配置成由至少一个处理器执行以下步骤：
获取目标对象的目标图像，其中，在获取目标图像前，目标对象上映射有识别标识，识别标识用于在目标图像中形成对应的识别特征；
根据识别特征按预设的映射关系确定目标图像对应的目标对象的属性，映射关系包括识别特征与目标对象的属性的对应关系。
其中，处理器的其他功能还可以参见上述实施例中记载的内容，这里不再一一 赘述。
基于同一发明构思，本说明书实施例还提供一种计算机可读存储介质，包括与电子设备结合使用的程序，程序可被处理器执行以完成以下步骤：
获取目标对象的目标图像，其中，在获取目标图像前，目标对象上映射有识别标识，识别标识用于在目标图像中形成对应的识别特征；
根据识别特征按预设的映射关系确定目标图像对应的目标对象的属性，映射关系包括识别特征与目标对象的属性的对应关系。
实施例4
图7为本说明书实施例提出的一种认证装置的结构示意图。
基于同一个申请构思，本说明书实施例记载的认证装置可以包括：
获取单元601，获取目标对象的第一目标图像，其中，在获取目标图像前，目标对象上映射有识别标识，识别标识用于在第一目标图像中形成对应的识别特征；
识别单元603，根据识别特征按预设的映射关系确定第一目标图像对应的目标对象的属性，映射关系包括识别特征与目标对象的属性的对应关系；
认证单元605，根据确定的结果，对目标对象进行认证。
可选地，所述的装置还包括：
在获取所述第一目标图像后，在预定时间段内获取所述目标对象的第二目标图像，所述第二目标图像为所述目标对象上未映射有所述识别标识的图像；
所述对所述目标对象进行认证包括：
按预设的数据处理策略，对所述第二目标图像进行数据处理；
根据数据处理的结果，确定所述目标对象的真伪。
基于同一发明构思，本说明书实施例还提供一种电子设备，包括至少一个处理器及存储器，所述存储器存储有程序，并且被配置成由至少一个处理器执行以下步骤：
获取目标对象的第一目标图像，其中，在获取目标图像前，目标对象上映射有识别标识，识别标识用于在第一目标图像中形成对应的识别特征；
根据识别特征按预设的映射关系确定第一目标图像对应的目标对象的属性，映射关系包括识别特征与目标对象的属性的对应关系；
根据确定的结果，对目标对象进行认证。
其中，处理器的其他功能还可以参见上述实施例中记载的内容，这里不再一一赘述。
基于同一发明构思，本说明书实施例还提供一种计算机可读存储介质，包括与电子设备结合使用的程序，程序可被处理器执行以完成以下步骤：
获取目标对象的第一目标图像，其中，在获取目标图像前，目标对象上映射有识别标识，识别标识用于在第一目标图像中形成对应的识别特征；
根据识别特征按预设的映射关系确定第一目标图像对应的目标对象的属性，映射关系包括识别特征与目标对象的属性的对应关系；
根据确定的结果，对目标对象进行认证。
在20世纪90年代，对于一个技术的改进可以很明显地区分是硬件上的改进(例如，对二极管、晶体管、开关等电路结构的改进)还是软件上的改进(对于方法流程的改进)。然而，随着技术的发展，当今的很多方法流程的改进已经可以视为硬件电路结构的直接改进。设计人员几乎都通过将改进的方法流程编程到硬件电路中来得到相应的硬件电路结构。因此，不能说一个方法流程的改进就不能用硬件实体模块来实现。例如，可编程逻辑器件(Programmable Logic Device,PLD)(例如现场可编程门阵列(Field Programmable Gate Array，FPGA))就是这样一种集成电路，其逻辑功能由用户对器件编程来确定。由设计人员自行编程来把一个数字系统“集成”在一片PLD上，而不需要请芯片制造厂商来设计和制作专用的集成电路芯片。而且，如今，取代手工地制作集成电路芯片，这种编程也多半改用“逻辑编译器(logic compiler)”软件来实现，它与程序开发撰写时所用的软件编译器相类似，而要编译之前的原始代码也得用特定的编程语言来撰写，此称之为硬件描述语言(Hardware Description Language，HDL)，而HDL也并非仅有一种，而是有许多种，如ABEL(Advanced Boolean Expression Language)、AHDL(Altera Hardware Description Language)、Confluence、CUPL(Cornell University Programming Language)、HDCal、JHDL(Java Hardware Description Language)、Lava、Lola、MyHDL、PALASM、RHDL(Ruby Hardware Description Language)等，目前最普遍使用的是VHDL(Very-High-Speed Integrated Circuit Hardware Description Language)与Verilog。本领域技术人员也应该清楚，只需要将方法流程用上述几种硬件描述语言稍作逻辑编程并编程到集成电路中，就可以很容易得到实现该逻辑方法流程的硬件电路。
控制器可以按任何适当的方式实现，例如，控制器可以采取例如微处理器或处理器以及存储可由该(微)处理器执行的计算机可读程序代码(例如软件或固件)的计算机可读介质、逻辑门、开关、专用集成电路(Application Specific Integrated Circuit，ASIC)、可编程逻辑控制器和嵌入微控制器的形式，控制器的例子包括但不限于以下微控制器：ARC 625D、Atmel AT91SAM、Microchip PIC18F26K20以及Silicone Labs C8051F320，存储器控制器还可以被实现为存储器的控制逻辑的一部分。本领域技术人员也知道，除了以纯计算机可读程序代码方式实现控制器以外，完全可以通过将方法步骤进行逻辑编程来使得控制器以逻辑门、开关、专用集成电路、可编程逻辑控制器和嵌入微控制器等的形式来实现相同功能。因此这种控制器可以被认为是一种硬件部件，而对其内包括的用于实现各种功能的装置也可以视为硬件部件内的结构。或者甚至，可以将用于实现各种功能的装置视为既可以是实现方法的软件模块又可以是硬件部件内的结构。
上述实施例阐明的系统、装置、模块或单元，具体可以由计算机芯片或实体实现，或者由具有某种功能的产品来实现。一种典型的实现设备为计算机。具体的，计算机例如可以为个人计算机、膝上型计算机、蜂窝电话、相机电话、智能电话、个人数字助理、媒体播放器、导航设备、电子邮件设备、游戏控制台、平板计算机、可穿戴设备或者这些设备中的任何设备的组合。
为了描述的方便，描述以上装置时以功能分为各种单元分别描述。当然，在实施本申请时可以把各单元的功能在同一个或多个软件和/或硬件中实现。
本领域内的技术人员应明白，本发明的实施例可提供为方法、系统、或计算机程序产品。因此，本发明可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。而且，本发明可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。
本发明是参照根据本发明实施例的方法、设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器，使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方 框中指定的功能的装置。
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上，使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。
在一个典型的配置中，计算设备包括一个或多个处理器(CPU)、输入/输出接口、网络接口和内存。
内存可能包括计算机可读介质中的非永久性存储器，随机存取存储器(RAM)和/或非易失性内存等形式，如只读存储器(ROM)或闪存(flash RAM)。内存是计算机可读介质的示例。
计算机可读介质包括永久性和非永久性、可移动和非可移动媒体可以由任何方法或技术来实现信息存储。信息可以是计算机可读指令、数据结构、程序的模块或其他数据。计算机的存储介质的例子包括，但不限于相变内存(PRAM)、静态随机存取存储器(SRAM)、动态随机存取存储器(DRAM)、其他类型的随机存取存储器(RAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、快闪记忆体或其他内存技术、只读光盘只读存储器(CD-ROM)、数字多功能光盘(DVD)或其他光学存储、磁盒式磁带，磁带磁磁盘存储或其他磁性存储设备或任何其他非传输介质，可用于存储可以被计算设备访问的信息。按照本文中的界定，计算机可读介质不包括暂存电脑可读媒体(transitory media)，如调制的数据信号和载波。
还需要说明的是，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、商品或者设备不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、商品或者设备所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括所述要素的过程、方法、商品或者设备中还存在另外的相同要素。
本申请可以在由计算机执行的计算机可执行指令的一般上下文中描述，例如程 序模块。一般地，程序模块包括执行特定任务或实现特定抽象数据类型的例程、程序、对象、组件、数据结构等等。也可以在分布式计算环境中实践本申请，在这些分布式计算环境中，由通过通信网络而被连接的远程处理设备来执行任务。在分布式计算环境中，程序模块可以位于包括存储设备在内的本地和远程计算机存储介质中。
本说明书中的各个实施例均采用递进的方式描述，各个实施例之间相同相似的部分互相参见即可，每个实施例重点说明的都是与其他实施例的不同之处。尤其，对于系统实施例而言，由于其基本相似于方法实施例，所以描述的比较简单，相关之处参见方法实施例的部分说明即可。
以上所述仅为本申请的实施例而已，并不用于限制本申请。对于本领域技术人员来说，本申请可以有各种更改和变化。凡在本申请的精神和原理之内所作的任何修改、等同替换、改进等，均应包含在本申请的权利要求范围之内。

Claims (22)

  1. 一种图像识别方法，包括：
    获取目标对象的目标图像，其中，在获取所述目标图像前，所述目标对象上映射有识别标识，所述识别标识用于在所述目标图像中形成对应的识别特征；
    根据所述识别特征按预设的映射关系确定所述目标图像对应的目标对象的属性，所述映射关系包括所述识别特征与所述目标对象的属性的对应关系。
  2. 根据权利要求1所述的方法，在所述获取目标对象的目标图像前，所述的方法还包括：检测所述目标对象，
    所述获取目标对象的目标图像包括：
    根据检测结果对应地获取所述目标对象的目标图像。
  3. 根据权利要求2所述的方法，所述检测所述目标对象包括：当所述目标对象上映射有所述识别标识时，基于所述识别标识检测所述目标对象。
  4. 根据权利要求2所述的方法，在所述检测所述目标对象中，所述的方法还包括：根据检测结果输出检测指示。
  5. 根据权利要求4所述的方法，所述检测指示包括语音播报和/或灯光指示。
  6. 根据权利要求2所述的方法，所述根据检测结果对应地获取所述目标对象的目标图像包括：检测到合格的所述目标对象时，在预设的采集时间内获取所述目标图像。
  7. 根据权利要求1所述的方法，所述识别标识包括包含定位特征的图案，所述定位特征用于所述图案的定位。
  8. 根据权利要求1所述的方法，所述方法用于移动终端中，所述移动终端包括显示单元和图像采集单元；
    所述获取目标对象的目标图像的步骤包括：
    通过所述图像采集单元获取目标对象的目标图像；
    在所述目标对象上映射识别标识的步骤包括：
    通过所述显示单元将在所述显示单元上呈现的所述识别标识映射在所述目标对象上；
    其中，所述图像采集单元与所述显示单元位于所述目标对象的同一侧。
  9. 根据权利要求8所述的方法，所述移动终端包括移动智能设备，所述显示单元包括屏幕，所述图像采集单元包括前置摄像头，
    所述通过所述图像采集单元获取所述目标图像包括：
    通过所述前置摄像头获取所述目标图像；
    所述通过所述显示单元将在所述显示单元上呈现的所述识别标识映射在所述目标对象上包括：
    通过所述屏幕将在所述屏幕上显示的所述识别标识映射在所述目标对象上。
  10. 根据权利要求1所述的方法，所述识别特征包括颜色特征、纹理特征、形状特征和空间关系特征之中的任意一种或多种。
  11. 一种认证方法，包括：
    获取目标对象的第一目标图像，其中，在获取所述第一目标图像前，所述目标对象上映射有识别标识，所述识别标识用于在所述第一目标图像中形成对应的识别特征；
    根据所述识别特征按预设的映射关系确定所述第一目标图像对应的目标对象的属性，所述映射关系包括所述识别特征与所述目标对象的属性的对应关系；
    根据确定的结果，对所述目标对象进行认证。
  12. 根据权利要求11所述的方法，所述方法还包括：
    在获取所述第一目标图像后，在预定时间段内获取所述目标对象的第二目标图像，所述第二目标图像为所述目标对象上未映射有所述识别标识的图像；
    所述对所述目标对象进行认证包括：
    按预设的数据处理策略，对所述第二目标图像进行数据处理；
    根据数据处理的结果，确定所述目标对象的真伪。
  13. 一种图像识别装置，包括：
    获取单元，获取目标对象的目标图像，其中，在获取所述目标图像前，所述目标对象上映射有识别标识，所述识别标识用于在所述目标图像中形成对应的识别特征；
    识别单元，根据所述识别特征按预设的映射关系确定所述目标图像对应的目标对象的属性，所述映射关系包括所述识别特征与所述目标对象的属性的对应关系。
  14. 根据权利要求13所述的装置，在获取所述目标图像前，所述装置还包括：
    检测单元，检测所述目标对象；
    所述获取目标对象的目标图像包括：
    根据检测结果对应地获取所述目标对象的目标图像。
  15. 根据权利要求14所述的装置，所述检测所述目标对象包括：当所述目标对象上映射有所述识别标识时，基于所述识别标识检测所述目标对象。
  16. 根据权利要求14所述的装置，在所述检测所述目标对象中，所述装置还包括：输出单元，根据检测结果输出检测指示。
  17. 根据权利要求14所述的装置，所述根据检测结果对应地获取所述目标对象的目 标图像包括：检测到合格的所述目标对象时，在预设的采集时间内获取所述目标图像。
  18. 根据权利要求13所述的装置，所述装置用于移动终端中，所述移动终端包括显示单元和图像采集单元；
    所述获取所述目标对象的目标图像的步骤包括：
    通过所述图像采集单元获取所述目标图像；
    在所述目标对象上映射识别标识的步骤包括：
    通过所述显示单元将在所述显示单元上呈现的所述识别标识映射在所述目标对象上；
    其中，所述图像采集单元与所述显示单元位于所述目标对象的同一侧。
  19. 一种认证装置，包括：
    获取单元，获取目标对象的第一目标图像，其中，在获取所述第一目标图像前，所述目标对象上映射有识别标识，所述识别标识用于在所述第一目标图像中形成对应的识别特征；
    识别单元，根据所述识别特征按预设的映射关系确定所述第一目标图像对应的目标对象的属性，所述映射关系包括所述识别特征与所述目标对象的属性的对应关系；
    认证单元，根据确定的结果，对所述目标对象进行认证。
  20. 根据权利要求19所述的装置，所述方法还包括：
    在获取所述第一目标图像后，在预定时间段内获取所述目标对象的第二目标图像，所述第二目标图像为所述目标对象上未映射有所述识别标识的图像；
    所述对所述目标对象进行认证包括：
    按预设的数据处理策略，对所述第二目标图像进行数据处理；
    根据数据处理的结果，确定所述目标对象的真伪。
  21. 一种电子设备，包括：至少一个处理器及存储器，所述存储器存储有程序，并且被配置成由至少一个处理器执行以下步骤：
    获取目标对象的目标图像，其中，在获取所述目标图像前，所述目标对象上映射有识别标识，所述识别标识用于在所述目标图像中形成对应的识别特征；
    根据所述识别特征按预设的映射关系确定所述目标图像对应的目标对象的属性，所述映射关系包括所述识别特征与所述目标对象的属性的对应关系。
  22. 一种电子设备，包括：至少一个处理器及存储器，所述存储器存储有程序，并且被配置成由至少一个处理器执行以下步骤：
    获取目标对象的第一目标图像，其中，在获取所述第一目标图像前，所述目标对象 上映射有识别标识，所述识别标识用于在所述第一目标图像中形成对应的识别特征；
    根据所述识别特征按预设的映射关系确定所述第一目标图像对应的目标对象的属性，所述映射关系包括所述识别特征与所述目标对象的属性的对应关系；
    根据确定的结果，对所述目标对象进行认证。
PCT/CN2020/070909 2019-05-29 2020-01-08 图像识别方法、装置、设备和认证方法、装置、设备 Ceased WO2020238232A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US16/805,530 US10796199B1 (en) 2019-05-29 2020-02-28 Image recognition and authentication

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910456505.1 2019-05-29
CN201910456505.1A CN110263775B (zh) 2019-05-29 2019-05-29 图像识别方法、装置、设备和认证方法、装置、设备

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US16/805,530 Continuation US10796199B1 (en) 2019-05-29 2020-02-28 Image recognition and authentication

Publications (1)

Publication Number Publication Date
WO2020238232A1 true WO2020238232A1 (zh) 2020-12-03

Family

ID=67915877

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/070909 Ceased WO2020238232A1 (zh) 2019-05-29 2020-01-08 图像识别方法、装置、设备和认证方法、装置、设备

Country Status (2)

Country Link
CN (1) CN110263775B (zh)
WO (1) WO2020238232A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113436246A (zh) * 2021-06-25 2021-09-24 中国工商银行股份有限公司 机柜内部空间检测方法、装置及系统

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110263775B (zh) * 2019-05-29 2021-03-05 创新先进技术有限公司 图像识别方法、装置、设备和认证方法、装置、设备
US10796199B1 (en) 2019-05-29 2020-10-06 Alibaba Group Holding Limited Image recognition and authentication
CN111178398B (zh) * 2019-12-13 2023-08-22 天翼商业保理有限公司 检测身份证图像信息篡改的方法、系统、存储介质及装置
CN111553285B (zh) * 2020-04-29 2022-10-04 上海智勘科技有限公司 在车辆远程定损过程中保护用户隐私信息的方法及系统
CN113256583A (zh) * 2021-05-24 2021-08-13 北京百度网讯科技有限公司 图像质量检测方法及装置、计算机设备和介质
CN113411477B (zh) * 2021-06-10 2023-03-10 支付宝(杭州)信息技术有限公司 一种图像的采集方法、装置及设备
CN114841955A (zh) * 2022-04-28 2022-08-02 杭州睿胜软件有限公司 一种生物物种的识别方法、装置、设备及存储介质
CN117407562B (zh) * 2023-12-13 2024-04-05 杭州海康威视数字技术股份有限公司 一种图像识别方法、系统及电子设备
CN119908709B (zh) * 2025-04-03 2025-06-24 内蒙古农业大学 一种非接触式牛瘤胃充盈度检测装置及检测方法

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030235334A1 (en) * 2002-06-19 2003-12-25 Pfu Limited Method for recognizing image
CN104252473A (zh) * 2013-06-27 2014-12-31 镇江睿泰信息科技有限公司 一种图像识别方法
CN105118048A (zh) * 2015-07-17 2015-12-02 北京旷视科技有限公司 翻拍证件图片的识别方法及装置
CN110263775A (zh) * 2019-05-29 2019-09-20 阿里巴巴集团控股有限公司 图像识别方法、装置、设备和认证方法、装置、设备

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130300101A1 (en) * 2012-05-11 2013-11-14 Document Security Systems, Inc. Laminated Documents and Cards Including Embedded Security Features
US10872488B2 (en) * 2016-06-03 2020-12-22 Veridas Digital Authentication Solutions, S.L. Method for authenticating a document
CN108573202A (zh) * 2017-03-17 2018-09-25 北京旷视科技有限公司 身份认证方法、装置和系统以及终端、服务器和存储介质
CN107944339B (zh) * 2017-10-20 2020-01-21 阿里巴巴集团控股有限公司 一种证件验证、身份验证方法和装置
CN108959862A (zh) * 2018-07-11 2018-12-07 南京环印防伪科技有限公司 一种基于数字水印和智能手机的证照防伪方法
CN109145768A (zh) * 2018-07-31 2019-01-04 北京旷视科技有限公司 得到带人脸属性的人脸数据的方法及装置
CN109409245A (zh) * 2018-09-30 2019-03-01 江苏满运软件科技有限公司 身份核验方法、系统、电子设备和存储介质

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030235334A1 (en) * 2002-06-19 2003-12-25 Pfu Limited Method for recognizing image
CN104252473A (zh) * 2013-06-27 2014-12-31 镇江睿泰信息科技有限公司 一种图像识别方法
CN105118048A (zh) * 2015-07-17 2015-12-02 北京旷视科技有限公司 翻拍证件图片的识别方法及装置
CN110263775A (zh) * 2019-05-29 2019-09-20 阿里巴巴集团控股有限公司 图像识别方法、装置、设备和认证方法、装置、设备

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113436246A (zh) * 2021-06-25 2021-09-24 中国工商银行股份有限公司 机柜内部空间检测方法、装置及系统

Also Published As

Publication number Publication date
CN110263775A (zh) 2019-09-20
CN110263775B (zh) 2021-03-05

Similar Documents

Publication Publication Date Title
WO2020238232A1 (zh) 图像识别方法、装置、设备和认证方法、装置、设备
AU2018350527B2 (en) Method and apparatus for verifying documents and identity
TWI718403B (zh) 一種證件驗證、身份驗證方法和裝置
CN111652223B (zh) 证件识别方法及装置
WO2019100814A1 (zh) 辅助物品的图像合规的方法、装置和电子设备
CN110263805B (zh) 证件验证、身份验证方法、装置及设备
CN110490200B (zh) 一种证件扫描方法、装置及设备
CN111310634B (zh) 证件类型识别模板的生成方法、证件识别方法及装置
CN115439851A (zh) 一种验证待识别证件图像的方法、系统及设备
US11392679B2 (en) Certificate verification
CN110705350B (zh) 证件识别方法及装置
CN111199231B (zh) 一种图像识别方法及装置
US10796199B1 (en) Image recognition and authentication
CN113705486B (zh) 检测证件真伪的方法及装置
CN112434687B (zh) 图片检测方法、装置、设备及存储介质
CN116560540A (zh) 一种对证件图像进行识别的方法、装置、设备和可读介质
HK40062631A (zh) 检测证件真伪的方法及装置
HK40062631B (zh) 检测证件真伪的方法及装置
CN119359979A (zh) 证件检测方法、装置、计算机可读存储介质、电子设备及计算机程序产品
CN118865203A (zh) 用于证件防伪检测的方法、装置、存储介质及电子设备

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20815379

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20815379

Country of ref document: EP

Kind code of ref document: A1