WO2018010652A1 - 图像识别回调通知的方法、服务器和计算机可读存储介质 - Google Patents

图像识别回调通知的方法、服务器和计算机可读存储介质 Download PDF

Info

Publication number
WO2018010652A1
WO2018010652A1 PCT/CN2017/092557 CN2017092557W WO2018010652A1 WO 2018010652 A1 WO2018010652 A1 WO 2018010652A1 CN 2017092557 W CN2017092557 W CN 2017092557W WO 2018010652 A1 WO2018010652 A1 WO 2018010652A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
callback
recognition result
requester
recognition
Prior art date
Application number
PCT/CN2017/092557
Other languages
English (en)
French (fr)
Inventor
王淼
陈会娟
刘冬冬
Original Assignee
腾讯科技（深圳）有限公司
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 腾讯科技（深圳）有限公司 filed Critical 腾讯科技（深圳）有限公司
Publication of WO2018010652A1 publication Critical patent/WO2018010652A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • 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
    • G06F18/22—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
    • G06V10/94—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
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10004—Still image; Photographic image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30196—Human being; Person

Definitions

  • the present invention relates to the field of image processing, and in particular, to a method, server and computer readable storage medium for image recognition callback notification.
  • the platform administrator As the requesting party, sends an image or the like to the image recognition system, and the image recognition system identifies whether the image contains certain specific content, such as containing pornographic information, and notifies the requesting party of the recognition result by a callback method.
  • the requester will receive all the recognition results, so that after receiving the recognition result, it needs to analyze and judge according to the information returned by the image recognition system, in order to get the required content, and a large number of invalid callback interfaces are called. This caused a waste of callback resources and caused excessive pressure on the callback identification server.
  • a method, server, and computer readable storage medium for image recognition callback notification are provided, which can save callback resources.
  • a method for image recognition callback notification comprising:
  • a server includes a memory and a processor, the memory storing computer readable instructions, the instructions being executed by the processor, causing the processor to perform the following steps:
  • One or more non-volatile readable storage media storing computer-executable instructions, when executed by one or more processors, cause the one or more processors to perform the following steps:
  • FIG. 1 is a schematic diagram of an application environment of a method for image recognition callback notification in an embodiment
  • FIG. 2A is a schematic diagram showing the internal structure of a terminal in an embodiment
  • 2B is a schematic diagram showing the internal structure of an identification server in an embodiment
  • FIG. 3 is a flow chart of a method for image recognition callback notification in an embodiment
  • FIG. 4 is a flow chart of a method for image recognition callback notification in another embodiment
  • FIG. 5 is a structural block diagram of an apparatus for image recognition callback notification in an embodiment
  • FIG. 6 is a structural block diagram of an apparatus for image recognition callback notification in another embodiment
  • FIG. 7 is a schematic diagram of the architecture of erotic image recognition in one embodiment.
  • FIG. 1 is a schematic diagram of an application environment of a method for image recognition callback notification in an embodiment.
  • the application environment includes a terminal 110 and an identification server 120.
  • the terminal 110 acquires the image sent by the requesting party and the requestor identifier, and uploads the acquired image and the requester identifier to the identification server 120.
  • the identification server 120 identifies the image to obtain the recognition result, and determines whether to perform the callback according to the first callback policy preset by the requester. If the callback is performed, the recognition result is sent to the requester, and only the first callback policy is met.
  • the recognition result of the image is recalled, which avoids calling back all the images and saves callback resources.
  • the terminal includes a processor, a non-volatile storage medium, an internal memory, a network interface, a display screen, and an input device connected by a system bus.
  • the non-volatile storage medium of the terminal stores an operating system, and further includes computer readable instructions that are executed by the processor to implement an image recognition callback notification method.
  • the processor is used to provide computing and control capabilities to support the operation of the entire terminal.
  • the internal memory in the terminal provides an environment for the operation of the device for recognizing the callback notification in the non-volatile storage medium, and the internal memory can store computer readable instructions, the computer readable instructions being When the processor is executed, the processor can be caused to perform a method of image recognition callback notification.
  • the network interface is used for network communication with the identification server, and the like.
  • the display screen of the terminal may be a liquid crystal display or an electronic ink display screen.
  • the input device may be a touch layer covered on the display screen, or may be a button, a trackball or a touchpad provided on the terminal housing, or may be an external device. Keyboard, trackpad or mouse.
  • the terminal can be a mobile phone, a tablet or a personal digital assistant or a wearable device.
  • FIG. 2A is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the terminal to which the solution of the present application is applied.
  • the specific terminal may include a ratio. More or fewer components are shown in the figures, or some components are combined, or have different component arrangements.
  • the identification server includes a processor, a non-volatile storage medium, an internal memory, and a network interface connected by a system bus.
  • the non-volatile storage medium of the identification server stores an operating system, a database, and computer readable instructions.
  • the database stores an identification method and a callback policy, and the computer readable instructions are executed by the processor to implement the identification server.
  • the processor of the identification server is used to provide computing and control capabilities to support the operation of the entire identification server.
  • the internal memory of the identification server is an environment providing an operation of a device for recognizing a callback notification in a non-volatile storage medium, wherein the internal memory can store computer readable instructions that are executed by the processor
  • the processor can be caused to perform a method of image recognition callback notification.
  • the network interface of the identification server is used to communicate with an external terminal via a network connection or the like.
  • the identification server can be implemented by a separate identification server or a cluster of identification servers composed of multiple identification servers. Those skilled in the art can understand that the structure shown in FIG. 2B is only a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation of the identification server to which the solution of the present application is applied.
  • the specific identification server may It includes more or fewer components than those shown in the figures, or some components are combined, or have different component arrangements.
  • FIG. 3 is a flow chart of a method of image recognition callback notification in one embodiment.
  • a method for image recognition callback notification which is run on the identification server in FIG. 1, includes:
  • Step 302 Acquire a requester identifier sent by the requesting party and a corresponding image recognition request.
  • the requesting party refers to a user who initiates an image recognition request, and the user may be a company or an administrator who provides a third-party application or a user who uses a third-party application.
  • the requester identity is used to uniquely identify the identity of the requestor.
  • the requester identification may be an instant messaging account, an email address, an identification number, an organization code, or an account registered on an identification server that identifies the image.
  • the registered account can be an individual user account or a corporate account.
  • the account registered on the identification server that identifies the image may include one or more of letters, numbers, and characters, and the like.
  • a direct identification interface is provided on the identification server, and the requester identifier sent by the requesting party and the corresponding image recognition request are obtained through the direct identification interface.
  • Step 304 Identify an image according to the image recognition request, and obtain a recognition result.
  • the identification server identifies the image in the image recognition request according to the image recognition request, and obtains the recognition result.
  • the identification rule may be configured in advance on the identification server, and the image may be identified according to the configured identification rule to obtain a recognition result.
  • the recognition result may include including specified content or not including specified content. Specify content as content that needs to be set as needed, such as pornography or a person or a piece of information.
  • the image recognition can be performed using an erotic image recognition algorithm based on the skin color feature.
  • the covariance matrix is:
  • the vector X (Cb, Cr) is used to represent a pixel in the image, and the probability that the pixel is skin color in the skin color distribution space can be represented by P(X).
  • Performing skin color segmentation on the acquired image first converting the image into an image in the YCbCr space, and then calculating a skin color probability of the pixel for each pixel of the image. If the skin color probability is not less than a set threshold, the pixel is a skin color pixel. It is indicated by 1, otherwise the pixel is not a skin color pixel, it is represented by 0, and finally the part of the skin color area is too small to be filtered out.
  • the ratio of the area of the skin color area to the entire image is obtained. If the ratio exceeds the preset ratio, the image is considered to be an erotic image. If the preset ratio is not exceeded, the image is not considered to be an erotic image.
  • the image to be identified is obtained, and the feature data corresponding to the image of the human body sensitive part matching the image to be identified in the pre-stored human sensitive part index is searched; and the estimated value corresponding to the image to be identified is calculated according to the matched feature data. And determining, according to the estimated value corresponding to the image to be identified, whether the image to be identified is an erotic image.
  • Step 306 Acquire a first callback policy corresponding to the requestor identifier.
  • the method further includes: the identification server acquires a callback policy configuration request sent by the requesting party, returns a policy configuration interface according to the callback policy configuration request, and obtains a first callback policy input by the requesting party on the policy configuration interface, and obtains the requester identifier. And the first callback policy, and the requestor identifier is stored corresponding to the first callback policy.
  • the first callback policy is set by the requesting party according to the need. If the recognition result of the image is an estimated value, the first callback policy may be within the first preset score range, and then the callback notification is performed, if not Within the preset score range, no callback notification is made. If the first callback strategy is that the estimated score is within the second preset score range, a callback notification is performed.
  • the first preset score range is smaller than the second preset score range. In the case of the porn score, if the image is viewed by a child or the like, if the score is within the first preset score range, a callback notification is made. For example, the first preset score range is [30,100], and the pornographic score range is [0,100].
  • the image When the image contains erotic scores above 30, the image is considered to contain pornography. content. If the object to be viewed is an adult, if the estimated value is within the second preset score range, a callback notification is performed, for example, the second preset score range is [50, 100], and the image contains pornography. When the score is above 50, the image is considered to contain pornographic content.
  • the first callback policy may include performing a callback of the recognition result of the preset number of images from the high to the low in the preset time range. For example, there are 100 images in 10 minutes for recognition, the scores of 100 images are sorted from high to low, and then the recognition results of 10 images are selected according to the evaluation value from high to low for callback notification.
  • Step 308 Determine whether to call back the recognition result according to the first callback policy, and if the callback is performed, notify the requester of the recognition result of the image.
  • the identification server determines whether to call back the recognition result according to the first callback policy, and if the result of the determination is that the callback is performed, the callback of the recognition result of the image is notified to the requesting party, and if the result of the determination is that no callback is performed, the identification server does not need to Make a callback notification.
  • the identification server callback notification requesting party may send the identification result to the terminal where the requesting party identifier is located, or send an email through the email address reserved by the requesting party, or send a message through the instant messaging account reserved by the requesting party, or The short message is sent by the mobile communication number reserved by the requesting party for notification, and the like.
  • the URL of the image (Uniform Resource Locator) can also be sent to the requesting party.
  • the method for recognizing the callback notification of the image after acquiring the image recognition request, identifying the image to obtain the recognition result, acquiring the first callback policy corresponding to the requester identifier, determining whether to perform the callback according to the first callback policy, and determining If a callback is needed, the callback result of the image is notified to the requester, and the callback is implemented according to the callback policy, which avoids callback of all images and saves callback resources.
  • the image is identified according to the image recognition request, and obtaining the recognition result includes: identifying the image according to the image recognition request, and obtaining a recognition result that the image includes an estimated value of the specified content.
  • the specified content can be set according to the requester's request, such as pornographic content or a certain person or a particular landscape.
  • the first callback policy performs a callback notification when the estimated value is within the first preset score range.
  • Determining, according to the first callback policy, whether to perform the callback of the recognition result, and if the callback is performed, notifying the requester of the recognition result of the image to the requester includes: determining whether the estimated value corresponding to the image is in the first preset score Within the range of values; if the estimated value is within the first predetermined score range, the recognition result of the image is notified to the requester.
  • the first preset score range is set as needed, such as [40, 100] or [50, 100].
  • the method for the image recognition callback notification further includes:
  • Step 402 Receive a requester identifier and a corresponding image uploaded by the requesting party, and store the requestor identifier and the corresponding image.
  • the identification server provides an image uploading interface, receives the requestor identifier uploaded by the requesting party and the corresponding image through the image uploading interface, and stores the image in the image storage system.
  • the image upload interface can be an interface for uploading images to third-party applications.
  • the requester refers to a user who initiates an image recognition request, which may be a company or an administrator who provides a third-party application or a user who uses a third-party application, and the like.
  • the requester identity is used to uniquely identify the identity of the requestor.
  • the requester identification may be an instant messaging account, an email address, an identification number, an organization code, or an account registered on an identification server that identifies the image.
  • the registered account can be an individual user account or a corporate account.
  • the account registered on the identification server that identifies the image may include one or more of letters, numbers, and characters, and the like.
  • Step 404 identifying the stored image to obtain a recognition result.
  • the identification server recognizes the stored image asynchronously by calling the identification service to obtain a recognition result.
  • Step 406 Acquire a second callback policy corresponding to the requester identifier.
  • the second callback policy may be set by the requesting party according to requirements or set by default.
  • the second callback strategy may include performing a callback notification when the estimated value is within the second preset score range, or performing a callback of the recognition result of the preset number of images from the high to the low within the preset time range, or
  • the method includes: blocking or deleting an image whose estimated value is within a third preset score range, or performing a callback notification when the estimated value is within the second preset score range and the estimated value is at a third preset score Images within the scope are banned or deleted.
  • Step 408 Determine whether to call back the recognition result of the stored image according to the second callback policy, and if the callback is performed, notify the requester of the recognition result of the stored image.
  • the identification server determines whether to call back the recognition result according to the second callback policy. If the result of the determination is that the callback is performed, the callback of the recognition result of the image is notified to the requesting party. If the result of the determination is that no callback is performed, the identification server does not need to perform the callback. Make a callback notification.
  • the identification server callback notification requesting party may send the identification result to the terminal where the requesting party identifier is located, or send an email through the email address reserved by the requesting party, or send a message through the instant messaging account reserved by the requesting party, or The short message is sent by the mobile communication number reserved by the requesting party for notification, and the like.
  • the URL of the image can also be sent to the requesting party.
  • the method for recognizing the callback notification is to obtain the uploaded image and store it, identify the stored image to obtain the recognition result, obtain a second callback policy corresponding to the requester identifier, and determine whether to call back the recognition result according to the second callback policy. If it is judged that the callback needs to be performed, the callback of the recognition result of the image is notified to the requester, the callback is performed according to the callback policy, the callback is avoided for all the images, the callback resource can be saved, and the upload image interface is provided, which is convenient to pass.
  • the upload image interface uploads images for storage.
  • the stored image is identified, and obtaining the recognition result includes: identifying the stored image to obtain a recognition result that the image includes an estimated value of the specified content.
  • the specified content can be set according to the requester's request, such as pornographic content or a certain person or a particular landscape.
  • the second callback strategy includes performing a callback notification when the estimated value is within the second predetermined score range.
  • the second preset score range is set as needed, such as [50, 100] or [60, 100].
  • the stored image is identified, and obtaining the recognition result includes: identifying the stored image to obtain a recognition result that the image includes an estimated value of the specified content.
  • the second callback strategy includes blocking or deleting images whose scores are within the third predetermined score range.
  • image blocking refers to marking an image as a prohibited image and cannot be displayed on the terminal.
  • Deleting means deleting the stored image and no longer displaying it on the terminal.
  • the third preset score range can be set or identify the default score range configured by the server as needed.
  • the third preset score range may be the same as or different from or partially the same as the second preset score range.
  • the method for recognizing the callback notification of the image further includes: determining whether the estimated value corresponding to the stored image is within a third preset score range; if the estimated value of the stored image is within the third preset score range , the stored image is banned or deleted, the banned result or the deleted result is obtained, and the banned result or the deleted result callback is notified to the requesting party.
  • the banned result can be marked as illegal and cannot be displayed.
  • the result of the deletion is that the notification image has been deleted.
  • the method for recognizing the callback notification of the image further includes: determining whether the estimated value corresponding to the stored image is within a third preset score range or within a second preset score range; if the storing If the estimated value of the image is within the third preset score range, the stored image is banned or deleted, the banned result or the deleted result is obtained, and the banned result or the deleted result callback is notified to the requester, the storage
  • the estimated value of the image is within the second preset score range, and the caller of the recognition result of the image is notified to the requester.
  • the banned result can be marked as illegal and cannot be displayed.
  • the result of the deletion is that the notification image has been deleted.
  • the above method for image recognition callback notification further comprises: transmitting a URL of the image to the requesting party.
  • the method for recognizing a callback notification of the image further includes: when the acquisition requester sends an image recognition request including an image that does not return the recognition result within a preset time, the acquisition requester transmits the image including the image again.
  • the request is identified, and the image recognition request including the image is acquired again without exceeding the preset number of retries.
  • the number of preset retries can be set according to the situation, such as 3 or 5 times.
  • FIG. 5 is a structural block diagram of an apparatus for image recognition callback notification in an embodiment.
  • an apparatus for image recognition callback notification includes an image recognition interface module 502, an image recognition processing module 504, a policy acquisition module 506, and a callback notification module 508. among them:
  • the image recognition interface module 502 is configured to obtain a requester identifier sent by the requesting party and a corresponding image recognition request.
  • the requesting party refers to a user who initiates an image recognition request, and the user may be a company or an administrator who provides a third-party application or a user who uses a third-party application.
  • the requester identity is used to uniquely identify the identity of the requestor.
  • the requester identification may be an instant messaging account, an email address, an identification number, an organization code, or an account registered on an identification server that identifies the image.
  • the registered account can be an individual user account or a corporate account.
  • the account registered on the identification server that identifies the image may include one or more of letters, numbers, and characters, and the like.
  • An image recognition interface is provided on the identification server, and the requester identifier sent by the requester and the corresponding image recognition request are acquired through the image recognition interface.
  • the image recognition processing module 504 is configured to identify the image according to the image recognition request, and obtain a recognition result.
  • the identification server identifies the image in the image recognition request according to the image recognition request, and obtains the recognition result.
  • the identification rule may be configured in advance on the identification server, and the image may be identified according to the configured identification rule to obtain a recognition result.
  • the recognition result may include including specified content or not including specified content. Specify content as content that needs to be set as needed, such as pornography or a person or a piece of information.
  • the policy obtaining module 506 is configured to obtain a first callback policy corresponding to the requestor identifier.
  • the first callback policy is set by the requesting party according to the need, for example, if the recognition result of the image is an estimated value, the first callback policy may be that the estimated value is within the first preset score range, and then the callback notification is performed, if not Within the default score range, no callback notifications are made. If the first callback strategy is that the estimated score is within the second preset score range, a callback notification is performed.
  • the first preset score range is smaller than the second preset score range. In the case of the porn score, if the image is viewed by a child or the like, if the score is within the first preset score range, a callback notification is made.
  • the first preset score range is [30,100]
  • the pornographic score range is [0,100].
  • the image contains erotic scores above 30, the image is considered to contain pornography. content.
  • the second preset score range is [50, 100]
  • the image contains pornography.
  • the score is above 50, the image is considered to contain pornographic content.
  • the first callback policy may include performing a callback of the recognition result of the preset number of images from the high to the low in the preset time range. For example, there are 100 images in 10 minutes for recognition, the scores of 100 images are sorted from high to low, and then the recognition results of 10 images are selected according to the evaluation value from high to low for callback notification.
  • the image recognition processing module 504 is further configured to determine, according to the first callback policy, whether to perform the callback of the recognition result.
  • the image recognition processing module 504 of the identification server determines whether to call back the recognition result according to the first callback policy. If the result of the determination is a callback, the image recognition result is notified to the requester, and if the determination result is no Callbacks do not require a callback notification.
  • the callback notification module 508 is configured to notify the requester of the recognition result of the image if the callback is performed.
  • the device for recognizing the callback notification After obtaining the image recognition request, the device for recognizing the callback notification obtains the recognition result, obtains the recognition result, acquires the first callback policy corresponding to the requester identifier, and determines whether to perform the callback according to the first callback policy, if it is determined If a callback is needed, the callback result of the image is notified to the requester, and the callback is implemented according to the callback policy, which avoids callback of all images and saves callback resources.
  • the first callback policy performs a callback notification when the estimated value is within the first preset score range
  • the image recognition processing module 504 is further configured to: identify an image according to the image recognition request, obtain a recognition result that the image includes an estimated value of the specified content, and determine whether the estimated value corresponding to the image is at the first preset score Within the scope.
  • the callback notification module 508 is further configured to notify the requester of the recognition result callback of the image if the estimated value is within the first preset score range.
  • FIG. 6 is a structural block diagram of an apparatus for image recognition callback notification in another embodiment.
  • an apparatus for image recognition callback notification includes a storage module 510 and a configuration module 512 in addition to an image recognition interface module 502, an image recognition processing module 504, a policy acquisition module 506, and a callback notification module 508. among them:
  • the storage module 510 is configured to receive the requester identifier and the corresponding image uploaded by the requesting party, and store the requestor identifier and the corresponding image.
  • an image uploading interface is provided, and the requester identifier uploaded by the requesting party and the corresponding image are received through the image uploading interface, and the image is stored in the image storage system.
  • the image upload interface can be an interface for uploading images to third-party applications.
  • the configuration module 512 is configured to obtain a callback policy configuration request sent by the requester, and return a policy configuration interface according to the callback policy configuration request, and obtain a callback policy input by the requesting party on the policy configuration interface, obtain the requester identifier and the callback policy, and request the request.
  • the party ID is stored corresponding to the callback policy.
  • the callback policy can include a first callback policy and a second callback policy.
  • the image recognition processing module 504 is further configured to identify the stored image to obtain a recognition result.
  • the policy obtaining module 506 is further configured to obtain a second callback policy corresponding to the requestor identifier.
  • the second callback policy may be set by the requesting party according to requirements or set by default.
  • the second callback strategy may include performing a callback notification when the estimated value is within the second preset score range, or performing a callback of the recognition result of the preset number of images from the high to the low within the preset time range, or Including bans or deletes the image with the estimated value within the third preset score range, or includes the estimated value at
  • the callback notification is performed when the second preset score range is within the range and the image whose score is within the third preset score range is banned or deleted.
  • the image recognition processing module 504 is further configured to determine, according to the second callback policy, whether to perform a callback of the stored image.
  • the callback notification module 508 is further configured to notify the requester of the recognition result callback of the stored image if the callback is performed.
  • the image recognition processing module 504 of the identification server determines whether to call back the recognition result according to the second callback policy. If the result of the determination is a callback, the callback notification module 508 notifies the requester of the recognition result of the image, if If the result of the judgment is that no callback is performed, no callback notification is required.
  • the callback notification module 508 of the identification server may call back the notification requester, and may send the identification result to the terminal where the requester identifier is located, or send an email through the email address reserved by the requesting party, or send an instant messaging account reserved by the requesting party.
  • the message is notified, or a short message is sent through the mobile communication number reserved by the requesting party for notification, and the like.
  • the URL of the image can also be sent to the requesting party.
  • the device for recognizing the callback notification image acquires the uploaded image and stores it, identifies the stored image to obtain a recognition result, acquires a second callback policy corresponding to the requester identifier, and determines whether to call back the recognition result according to the second callback policy. If it is judged that the callback needs to be performed, the callback of the recognition result of the image is notified to the requester, the callback is performed according to the callback policy, the callback is avoided for all the images, the callback resource can be saved, and the upload image interface is provided, which is convenient to pass.
  • the upload image interface uploads images for storage.
  • the second callback strategy includes performing a callback notification when the estimated value is within the second predetermined score range.
  • the image recognition processing module 504 is further configured to identify the stored image, obtain a recognition result that the image includes an estimated value of the specified content, and determine whether the estimated value corresponding to the stored image is at a second preset score.
  • the callback notification module 508 is further configured to: if the estimated value of the stored image is within the second preset score range, notify the requester of the recognition result callback of the stored image.
  • the image recognition processing module 504 is further configured to identify the stored image to obtain a recognition result that the image includes an estimated value of the specified content.
  • the second callback strategy includes blocking or deleting an image whose estimated value is within a third preset score range.
  • the image recognition processing module 504 is further configured to determine whether the estimated value corresponding to the stored image is within a third preset score range, and if the estimated value of the stored image is within the third preset score range , the stored image is banned or deleted, and the banned result or the deleted result is obtained.
  • the callback notification module 508 is further configured to notify the requester of the banned result or the deletion result callback.
  • the banned result can be marked as illegal and cannot be displayed.
  • the result of the deletion is that the notification image has been deleted.
  • the image recognition interface module 502 acquires the requester to resend the image recognition request including the image, and again Obtaining an image recognition request containing the image does not exceed a preset number of retries.
  • the number of preset retries can be set according to the situation, such as 3 or 5 times.
  • the implementation process of the method and apparatus for image recognition callback notification is described below in conjunction with a specific application scenario.
  • the method for image recognition callback notification is applied to erotic image recognition as an example.
  • the architecture of erotic image recognition is as shown in FIG. 7 , and includes a callback policy configuration module 702, an erotic image recognition processing module 704, an erotic image recognition interface 706, and a callback notification module 708. And a storage module 710.
  • the callback policy configuration module 702 obtains data input by the user on the callback policy configuration interface, and generates a callback policy according to the input data.
  • the erotic image recognition interface 704 receives the erotic image recognition request transmitted by the requesting party, and transmits the image in the erotic image recognition request to the erotic image recognition processing module 704.
  • the erotic image recognition processing module 704 identifies the image, obtains the pornographic score of the image, and then reads the callback policy configured by the requester, and determines whether the pornographic score of the image is within the preset score range according to the callback policy. And the recognition result of the image is sent to the callback notification module 708.
  • the callback notification module 708 is configured to notify the requestor or the server configured by the requestor of the recognition result callback of the image.
  • the storage module 710 is configured to store an image uploaded by the requesting party through the third party application.
  • the erotic image recognition processing module 704 recognizes the image stored in the storage module 710, and obtains the pornographic score of the image.
  • the value is obtained by the requester's callback policy, and the image is banned or deleted according to the callback policy, and the corresponding banned result or the deleted result is obtained, and the banned result or the deleted result is sent to the callback notification module 708, and the image information that needs to be called back is obtained. Sended to callback notification module 708.
  • the callback notification module 708 notifies the requestor or the server configured by the requestor of the recognition result of the uniform resource locator of the image, the pornographic score, whether it is blocked or deleted. If the image recognition request sent by the requester fails, multiple attempts are made.
  • the storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), or the like.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Quality & Reliability (AREA)
  • General Engineering & Computer Science (AREA)
  • Library & Information Science (AREA)
  • Databases & Information Systems (AREA)
  • Artificial Intelligence (AREA)
  • Software Systems (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Multimedia (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Information Transfer Between Computers (AREA)
  • Image Analysis (AREA)

Abstract

一种图像识别回调通知的方法包括：获取请求方发送的请求方标识及对应的图像识别请求；根据所述图像识别请求对图像进行识别，得到识别结果；获取与所述请求方标识对应的第一回调策略；根据所述第一回调策略判断是否将所述识别结果进行回调，若进行回调，则将所述图像的识别结果回调通知给所述请求方。

Description

图像识别回调通知的方法、服务器和计算机可读存储介质
本申请要求于2016年7月12日提交中国专利局、申请号为201610548024.X、发明名称为“图像识别回调通知的方法和装置”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本发明涉及图像处理领域，特别是涉及一种图像识别回调通知的方法、服务器和计算机可读存储介质。
背景技术
随着计算机技术和网络技术的发展，越来越多的用户拍摄图像，并上传到各种平台进行分享，平台管理者需要对上传的图像进行识别。为此，平台管理者作为请求方，会将图像等发送给图像识别系统，图像识别系统识别该图像是否包含某些特定内容，如包含色情信息，并将识别结果通过回调方式通知给请求方。
通常，请求方会收到所有的识别结果，导致在收到识别结果后，还需要根据图像识别系统返回的信息，再进行分析和判断，才能得到所需要的内容，大量无效的调用了回调接口，造成了回调资源的浪费，对回调识别服务器造成过大压力。
发明内容
根据本申请的各种实施例，提供一种图像识别回调通知的方法、服务器和计算机可读存储介质，能节省回调资源。
一种图像识别回调通知的方法，包括：
获取请求方发送的请求方标识及对应的图像识别请求；
根据所述图像识别请求对图像进行识别，得到识别结果；
获取与所述请求方标识对应的第一回调策略；
根据所述第一回调策略判断是否将所述识别结果进行回调，若进行回调，则将所述图像的识别结果回调通知给所述请求方。
一种服务器，包括存储器及处理器，所述存储器中储存有计算机可读指令，所述指令被所述处理器执行时，使得所述处理器执行以下步骤：
获取请求方发送的请求方标识及对应的图像识别请求；
根据所述图像识别请求对图像进行识别，得到识别结果；
获取与所述请求方标识对应的第一回调策略；及
根据所述第一回调策略判断是否将所述识别结果进行回调，若进行回调，则将所述图像的识别结果回调通知给所述请求方。
一个或多个存储有计算机可执行指令的非易失性可读存储介质，所述计算机可执行指令被一个或多个处理器执行时，使得所述一个或多个处理器执行以下步骤：
获取请求方发送的请求方标识及对应的图像识别请求；
根据所述图像识别请求对图像进行识别，得到识别结果；
获取与所述请求方标识对应的第一回调策略；及
根据所述第一回调策略判断是否将所述识别结果进行回调，若进行回调，则将所述图像的识别结果回调通知给所述请求方。
本发明的一个或多个实施例的细节在下面的附图和描述中提出。本发明的其它特征、目的和优点将从说明书、附图以及权利要求书变得明显。
附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1为一个实施例中图像识别回调通知的方法的应用环境示意图；
图2A为一个实施例中终端的内部结构示意图；
图2B为一个实施例中识别服务器的内部结构示意图；
图3为一个实施例中图像识别回调通知的方法的流程图；
图4为另一个实施例中图像识别回调通知的方法的流程图；
图5为一个实施例中图像识别回调通知的装置的结构框图；
图6为另一个实施例中图像识别回调通知的装置的结构框图；
图7为一个实施例中色情图像识别的架构示意图。
具体实施方式
为了使本发明的目的、技术方案及优点更加清楚明白，以下结合附图及实施例，对本发明进行进一步详细说明。应当理解，此处所描述的具体实施例仅仅用以解释本发明，并不用于限定本发明。
图1为一个实施例中图像识别回调通知的方法的应用环境示意图。如图1所示，该应用环境包括终端110和识别服务器120。终端110获取请求方发送的图像及请求方标识，并将获取的图像及请求方标识上传到识别服务器120。识别服务器120识别该图像得到识别结果，并根据请求方预先设置的第一回调策略判断是否将识别结果进行回调，若进行回调，则将识别结果发送给请求方，只将符合第一回调策略的图像的识别结果进行回调，避免了将所有图像进行回调，能节省回调资源。
图2A为一个实施例中终端的内部结构示意图。如图2A所示，该终端包括通过系统总线连接的处理器、非易失性存储介质、内存储器、网络接口、显示屏和输入装置。其中，终端的非易失性存储介质存储有操作系统，还包括计算机可读指令，该计算机可读指令被处理器执行时以实现一种图像识别回调通知的方法。该处理器用于提供计算和控制能力，支撑整个终端的运行。终端中的内存储器为非易失性存储介质中的图像识别回调通知的装置的运行提供环境，该内存储器中可储存有计算机可读指令，该计算机可读指令被所 述处理器执行时，可使得所述处理器执行一种图像识别回调通知的方法。网络接口用于与识别服务器进行网络通信等。终端的显示屏可以是液晶显示屏或者电子墨水显示屏等，输入装置可以是显示屏上覆盖的触摸层，也可以是终端外壳上设置的按键、轨迹球或触控板，也可以是外接的键盘、触控板或鼠标等。该终端可以是手机、平板电脑或者个人数字助理或穿戴式设备等。本领域技术人员可以理解，图2A中示出的结构，仅仅是与本申请方案相关的部分结构的框图，并不构成对本申请方案所应用于其上的终端的限定，具体的终端可以包括比图中所示更多或更少的部件，或者组合某些部件，或者具有不同的部件布置。
图2B为一个实施例中识别服务器(或云端等)的内部结构示意图。如图2B所示，该识别服务器包括通过系统总线连接的处理器、非易失性存储介质、内存储器和网络接口。其中，该识别服务器的非易失性存储介质存储有操作系统、数据库和计算机可读指令，数据库中存储有识别方法和回调策略，该计算机可读指令被处理器执行时以实现适用于识别服务器的一种图像识别回调通知的方法。该识别服务器的处理器用于提供计算和控制能力，支撑整个识别服务器的运行。该识别服务器的内存储器为非易失性存储介质中的图像识别回调通知的装置的运行提供环境，该内存储器中可储存有计算机可读指令，该计算机可读指令被所述处理器执行时，可使得所述处理器执行一种图像识别回调通知的方法。该识别服务器的网络接口用于据以与外部的终端通过网络连接通信等。识别服务器可以用独立的识别服务器或者是多个识别服务器组成的识别服务器集群来实现。本领域技术人员可以理解，图2B中示出的结构，仅仅是与本申请方案相关的部分结构的框图，并不构成对本申请方案所应用于其上的识别服务器的限定，具体的识别服务器可以包括比图中所示更多或更少的部件，或者组合某些部件，或者具有不同的部件布置。
图3为一个实施例中图像识别回调通知的方法的流程图。如图3所示，一种图像识别回调通知的方法，运行于图1中的识别服务器上，包括：
步骤302，获取请求方发送的请求方标识及对应的图像识别请求。
本实施例中，请求方是指发起图像识别请求的用户，该用户可为提供第三方应用程序的企业或管理人员或使用第三方应用程序的用户等。
请求方标识是用于唯一标识请求方身份的。请求方标识可为即时通讯账号、电子邮箱、身份证号码、组织代码或在识别图像的识别服务器上注册的账户等。该注册的账户可为个人用户账户或企业账户等。在识别图像的识别服务器上注册的账户可包括字母、数字和字符中的一种或多种等。
在识别服务器上提供了直接识别接口，通过直接识别接口获取请求方发送的请求方标识及对应的图像识别请求。
步骤304，根据该图像识别请求对图像进行识别，得到识别结果。
本实施例中，识别服务器根据该图像识别请求对图像识别请求中的图像进行识别，得到识别结果。可预先在识别服务器上配置识别规则，根据配置的识别规则对图像进行识别，得到识别结果。该识别结果可包括包含指定内容或不包含指定内容等。指定内容为根据需要设置的内容，如色情内容或某个人物或某个信息等。
若是识别色情图像，则对图像进行识别可采用基于肤色特征的色情图像识别算法。首先建立肤色模型，基于常用的色度空间RGB(红、绿、蓝三原色)、HIS(色调、饱和度、亮度)、YIQ(NTSC制式的光亮度和色度模型)、YUV(PAL制式的光亮度和色度模型)和YCbCr(亮度分量、蓝色色度分量和红色色度分量模型)等。以YCbCr色度空间为例，选取大量的肤色样本进行统计，结果表明不同人种的肤色区别主要受亮度影响，受色度的影响较小，肤色在CbCr空间的分布呈现良好的聚类特性，且统计分布满足：77≤Cb≤127，133≤Cr≤173。这表明肤色在CbCr空间集中分布，符合高斯分布特征，采用高斯模型来描述。假设有N个肤色像素Xi(Cbi，Cri)(i＝1，2，…，N)作为训练样本，则均值向量
Figure PCTCN2017092557-appb-000001
其中：
Figure PCTCN2017092557-appb-000002
Figure PCTCN2017092557-appb-000003
协方差矩阵则为：
Figure PCTCN2017092557-appb-000004
用向量X(Cb，Cr)表示图像中的一个像素，则该像素在肤色分布空间中是肤色的概率可用P(X)表示。
Figure PCTCN2017092557-appb-000005
选取多幅不同年龄和性别人的彩色图像，用Photoshop作为工具手工分割出指定尺寸的肤色图像，计算出高斯模型的均值向量和协方差矩阵。
对获取的图像进行肤色区域分割，先将图像转换为YCbCr空间中的图像，再对图像的每一个像素计算像素的肤色概率，如果肤色概率不小于设定的阈值，则该像素是肤色像素，用1表示，否则该像素不是肤色像素，用0表示，最后把肤色区域面积过小的部分过滤掉。获取肤色区域面积占整幅图像的比例，若比例超过预设比例，则认为该图像为色情图像，若未超过预设比例，则不认为图像为色情图像。
在其他实施例中，获取待识别图像，查找预先存储的人体敏感部位索引中与待识别图像匹配的人体敏感部位图片所对应的特征数据；根据匹配的特征数据计算待识别图像对应的估分值；根据待识别图像对应的估分值判断待识别图像是否为色情图像。
步骤306，获取与该请求方标识对应的第一回调策略。
在步骤306之前，还包括：识别服务器获取请求方发送的回调策略配置请求，根据该回调策略配置请求返回策略配置界面，并获取请求方在策略配置界面输入的第一回调策略，获取请求方标识和第一回调策略，并将请求方标识与第一回调策略对应存储。
第一回调策略由请求方根据需要设置，如对图像的识别结果为一估分值，则第一回调策略可为估分值在第一预设分值范围内，则进行回调通知，若不 在预设分值范围内，则不进行回调通知。再如第一回调策略为估分值在第二预设分值范围内，则进行回调通知。第一预设分值范围小于第二预设分值范围。以色情分值为例，若图像被浏览的对象为儿童等场合，则估分值在第一预设分值范围内，则进行回调通知。例如第一预设分值范围为[30，100]，色情的估分值范围为[0，100]，当图像中包含的色情的估分值在30以上时，该图像即被认为含有色情内容。若图像被浏览的对象为成年人，则估分值在第二预设分值范围内，则进行回调通知，例如第二预设分值范围为[50，100]，图像中包含的色情的估分值在50以上时，该图像被认为含有色情内容。
在其他实施例中，第一回调策略可包括在预设时间范围内估分值从高到低中预设数量的图像的识别结果进行回调。例如在10分钟内有100张图像进行识别，100张图像的估分值从高到低进行排序，然后按照估分值从高到低选择10张图像的识别结果进行回调通知。
步骤308，根据该第一回调策略判断是否将该识别结果进行回调，若进行回调，则将该图像的识别结果回调通知给该请求方。
本实施例中，识别服务器根据第一回调策略判断是否对识别结果进行回调，若判断结果为进行回调，则将图像的识别结果回调通知给请求方，若判断结果为不进行回调，则不需要进行回调通知。
识别服务器回调通知请求方，可将识别结果发送给请求方标识所在的终端，或者通过请求方预留的电子邮箱发邮件进行通知，或者通过请求方预留的即时通讯账号发消息进行通知，或者通过请求方预留的移动通信号码发短消息进行通知等。此外，还可将图像的URL(Uniform Resource Locator，统一资源定位符)发送给请求方。
上述图像识别回调通知的方法，获取图像识别请求后，对图像进行识别得到识别结果，获取与请求方标识对应的第一回调策略，根据第一回调策略判断是否将识别结果进行回调，若判断出需要进行回调，则将图像的识别结果回调通知给请求方，实现了根据回调策略进行回调，避免了对全部图像进行回调，能节省回调资源。
在一个实施例中，该根据该图像识别请求对图像进行识别，得到识别结果包括：根据该图像识别请求对图像进行识别，得到该图像包含指定内容的估分值的识别结果。
该指定内容可根据请求方要求进行设置，如色情内容或某个人物或某个特定风景等。
该第一回调策略为估分值在第一预设分值范围内时进行回调通知。
该根据该第一回调策略判断是否将该识别结果进行回调，若进行回调，则将该图像的识别结果回调通知给该请求方包括：判断该图像对应的估分值是否在第一预设分值范围内；若该估分值在该第一预设分值范围内，则将该图像的识别结果回调通知给该请求方。
第一预设分值范围根据需要设定，如[40，100]或[50，100]等。
在一个实施例中，如图4所示，上述图像识别回调通知的方法还包括：
步骤402，接收请求方上传的请求方标识及对应的图像，并存储该请求方标识及对应的图像。
本实施例中，识别服务器提供图像上传接口，通过图像上传接口接收请求方上传的请求方标识及对应的图像，并将图像存储在图像存储系统中。图像上传接口可为开放给第三方应用上传图像的接口。
请求方是指发起图像识别请求的用户，该用户可为提供第三方应用程序的企业或管理人员或使用第三方应用程序的用户等。
请求方标识是用于唯一标识请求方身份的。请求方标识可为即时通讯账号、电子邮箱、身份证号码、组织代码或在识别图像的识别服务器上注册的账户等。该注册的账户可为个人用户账户或企业账户等。在识别图像的识别服务器上注册的账户可包括字母、数字和字符中的一种或多种等。
步骤404，对该存储的图像进行识别，得到识别结果。
本实施例中，识别服务器对存储的图像异步调用识别服务进行识别，得到识别结果。
步骤406，获取与该请求方标识对应的第二回调策略。
本实施例中，第二回调策略可为请求方根据需要设置或者默认设置。第二回调策略可包括估分值在第二预设分值范围内时进行回调通知，或者在预设时间范围内估分值从高到低中预设数量的图像的识别结果进行回调，或者包括将估分值在第三预设分值范围内的图像封禁或删除，或者包括估分值在第二预设分值范围内时进行回调通知且将估分值在第三预设分值范围内的图像封禁或删除等。
步骤408，根据该第二回调策略判断是否将该存储的图像的识别结果进行回调，若进行回调，则将该存储的图像的识别结果回调通知给该请求方。
本实施例中，识别服务器根据第二回调策略判断是否对识别结果进行回调，若判断结果为进行回调，则将图像的识别结果回调通知给请求方，若判断结果为不进行回调，则不需要进行回调通知。
识别服务器回调通知请求方，可将识别结果发送给请求方标识所在的终端，或者通过请求方预留的电子邮箱发邮件进行通知，或者通过请求方预留的即时通讯账号发消息进行通知，或者通过请求方预留的移动通信号码发短消息进行通知等。此外，还可将图像的URL发送给请求方。
上述图像识别回调通知的方法，获取上传的图像并进行存储，对存储的图像进行识别得到识别结果，获取与请求方标识对应的第二回调策略，根据第二回调策略判断是否将识别结果进行回调，若判断出需要进行回调，则将图像的识别结果回调通知给请求方，实现了根据回调策略进行回调，避免了对全部图像进行回调，能节省回调资源，且提供了上传图像接口，方便通过上传图像接口上传图像进行存储。
在一个实施例中，该对该存储的图像进行识别，得到识别结果包括：对该存储的图像进行识别，得到该图像包含指定内容的估分值的识别结果。
该指定内容可根据请求方要求进行设置，如色情内容或某个人物或某个特定风景等。
该第二回调策略包括估分值在第二预设分值范围内时进行回调通知。
第二预设分值范围根据需要设定，如[50，100]或[60，100]等。
该根据该第二回调策略判断是否将该存储的图像的识别结果进行回调，若进行回调，则将该存储的图像的识别结果回调通知给该请求方，包括：判断该存储的图像对应的估分值是否在第二预设分值范围内；若该存储的图像的估分值在该第二预设分值范围内，则将该存储的图像的识别结果回调通知给该请求方。
在一个实施例中，该对该存储的图像进行识别，得到识别结果包括：对该存储的图像进行识别，得到该图像包含指定内容的估分值的识别结果。
第二回调策略包括将估分值在第三预设分值范围内的图像封禁或删除。其中，图像封禁是指将图像标记为违禁图像且无法显示在终端上。删除是指将存储的图像删除，且不再在终端上显示。第三预设分值范围可根据需要设定或识别服务器配置的默认分值范围。第三预设分值范围可与第二预设分值范围相同或不同或部分相同。
上述图像识别回调通知的方法还包括：判断该存储的图像对应的估分值是否在第三预设分值范围内；若该存储的图像的估分值在该第三预设分值范围内，则将该存储的图像封禁或删除，得到封禁结果或删除结果，并将该封禁结果或删除结果回调通知该请求方。封禁结果可为标记为违法图像且无法显示。删除结果为通知图像已删除。
在一个实施例中，上述图像识别回调通知的方法还包括：判断该存储的图像对应的估分值是否在第三预设分值范围内或在第二预设分值范围内；若该存储的图像的估分值在该第三预设分值范围内，则将该存储的图像封禁或删除，得到封禁结果或删除结果，并将该封禁结果或删除结果回调通知该请求方，该存储的图像的估分值在第二预设分值范围内，则将图像的识别结果回调通知给请求方。封禁结果可为标记为违法图像且无法显示。删除结果为通知图像已删除。
在一个实施例中，上述图像识别回调通知的方法还包括：将图像的URL发送给请求方。
在一个实施例中，上述图像识别回调通知的方法还包括：当获取请求方发送包含某图像的图像识别请求在预设时间内未返回识别结果时，获取请求方再次发送包含所述图像的图像识别请求，且再次获取包含所述图像的图像识别请求不超过预设重试次数。预设重试次数可根据情况设定，如3次或5次等。
图5为一个实施例中图像识别回调通知的装置的结构框图。如图5所示，一种图像识别回调通知的装置，包括图像识别接口模块502、图像识别处理模块504、策略获取模块506和回调通知模块508。其中：
图像识别接口模块502用于获取请求方发送的请求方标识及对应的图像识别请求。
本实施例中，请求方是指发起图像识别请求的用户，该用户可为提供第三方应用程序的企业或管理人员或使用第三方应用程序的用户等。
请求方标识是用于唯一标识请求方身份的。请求方标识可为即时通讯账号、电子邮箱、身份证号码、组织代码或在识别图像的识别服务器上注册的账户等。该注册的账户可为个人用户账户或企业账户等。在识别图像的识别服务器上注册的账户可包括字母、数字和字符中的一种或多种等。
在识别服务器上提供了图像识别接口，通过图像识别接口获取请求方发送的请求方标识及对应的图像识别请求。
图像识别处理模块504用于根据该图像识别请求对图像进行识别，得到识别结果。
本实施例中，识别服务器根据该图像识别请求对图像识别请求中的图像进行识别，得到识别结果。可预先在识别服务器上配置识别规则，根据配置的识别规则对图像进行识别，得到识别结果。该识别结果可包括包含指定内容或不包含指定内容等。指定内容为根据需要设置的内容，如色情内容或某个人物或某个信息等。
策略获取模块506用于获取与该请求方标识对应的第一回调策略。
第一回调策略由请求方根据需要设置，如对图像的识别结果为一估分值，则第一回调策略可为估分值在第一预设分值范围内，则进行回调通知，若不在预设分值范围内，则不进行回调通知。再如第一回调策略为估分值在第二预设分值范围内，则进行回调通知。第一预设分值范围小于第二预设分值范围。以色情分值为例，若图像被浏览的对象为儿童等场合，则估分值在第一预设分值范围内，则进行回调通知。例如第一预设分值范围为[30，100]，色情的估分值范围为[0，100]，当图像中包含的色情的估分值在30以上时，该图像即被认为含有色情内容。若图像被浏览的对象为成年人，则估分值在第二预设分值范围内，则进行回调通知，例如第二预设分值范围为[50，100]，图像中包含的色情的估分值在50以上时，该图像被认为含有色情内容。
在其他实施例中，第一回调策略可包括在预设时间范围内估分值从高到低中预设数量的图像的识别结果进行回调。例如在10分钟内有100张图像进行识别，100张图像的估分值从高到低进行排序，然后按照估分值从高到低选择10张图像的识别结果进行回调通知。
该图像识别处理模块504还用于根据该第一回调策略判断是否将该识别结果进行回调。
本实施例中，识别服务器的图像识别处理模块504根据第一回调策略判断是否对识别结果进行回调，若判断结果为进行回调，则将图像的识别结果回调通知给请求方，若判断结果为不进行回调，则不需要进行回调通知。
回调通知模块508用于若进行回调，则将该图像的识别结果回调通知给该请求方。
上述图像识别回调通知的装置，获取图像识别请求后，对图像进行识别得到识别结果，获取与请求方标识对应的第一回调策略，根据第一回调策略判断是否将识别结果进行回调，若判断出需要进行回调，则将图像的识别结果回调通知给请求方，实现了根据回调策略进行回调，避免了对全部图像进行回调，能节省回调资源。
在一个实施例中，该第一回调策略为估分值在第一预设分值范围内时进行回调通知；
该图像识别处理模块504还用于根据该图像识别请求对图像进行识别，得到该图像包含指定内容的估分值的识别结果，以及判断该图像对应的估分值是否在第一预设分值范围内。
该回调通知模块508还用于若该估分值在该第一预设分值范围内，则将该图像的识别结果回调通知给该请求方。
图6为另一个实施例中图像识别回调通知的装置的结构框图。如图6所示，一种图像识别回调通知的装置，除了包括图像识别接口模块502、图像识别处理模块504、策略获取模块506和回调通知模块508，还包括存储模块510和配置模块512。其中：
存储模块510用于接收请求方上传的请求方标识及对应的图像，并存储该请求方标识及对应的图像。
本实施例中，提供图像上传接口，通过图像上传接口接收请求方上传的请求方标识及对应的图像，并将图像存储在图像存储系统中。图像上传接口可为开放给第三方应用上传图像的接口。
配置模块512用于获取请求方发送的回调策略配置请求，根据该回调策略配置请求返回策略配置界面，并获取请求方在策略配置界面输入的回调策略，获取请求方标识和回调策略，并将请求方标识与回调策略对应存储。回调策略可包括第一回调策略和第二回调策略。
在一个实施例中，该图像识别处理模块504还用于对该存储的图像进行识别，得到识别结果。
该策略获取模块506还用于获取与该请求方标识对应的第二回调策略。
本实施例中，第二回调策略可为请求方根据需要设置或者默认设置。第二回调策略可包括估分值在第二预设分值范围内时进行回调通知，或者在预设时间范围内估分值从高到低中预设数量的图像的识别结果进行回调，或者包括将估分值在第三预设分值范围内的图像封禁或删除，或者包括估分值在 第二预设分值范围内时进行回调通知且将估分值在第三预设分值范围内的图像封禁或删除等。
该图像识别处理模块504还用于根据该第二回调策略判断是否将该存储的图像的识别结果进行回调。
该回调通知模块508还用于若进行回调，则将该存储的图像的识别结果回调通知给该请求方。
本实施例中，识别服务器的图像识别处理模块504根据第二回调策略判断是否对识别结果进行回调，若判断结果为进行回调，则回调通知模块508将图像的识别结果回调通知给请求方，若判断结果为不进行回调，则不需要进行回调通知。
识别服务器的回调通知模块508回调通知请求方，可将识别结果发送给请求方标识所在的终端，或者通过请求方预留的电子邮箱发邮件进行通知，或者通过请求方预留的即时通讯账号发消息进行通知，或者通过请求方预留的移动通信号码发短消息进行通知等。此外，还可将图像的URL发送给请求方。
上述图像识别回调通知的装置，获取上传的图像并进行存储，对存储的图像进行识别得到识别结果，获取与请求方标识对应的第二回调策略，根据第二回调策略判断是否将识别结果进行回调，若判断出需要进行回调，则将图像的识别结果回调通知给请求方，实现了根据回调策略进行回调，避免了对全部图像进行回调，能节省回调资源，且提供了上传图像接口，方便通过上传图像接口上传图像进行存储。
该第二回调策略包括估分值在第二预设分值范围内时进行回调通知。
该图像识别处理模块504还用于对该存储的图像进行识别，得到该图像包含指定内容的估分值的识别结果，以及判断该存储的图像对应的估分值是否在第二预设分值范围内。
该回调通知模块508还用于若该存储的图像的估分值在该第二预设分值范围内，则将该存储的图像的识别结果回调通知给该请求方。
在一个实施例中，该图像识别处理模块504还用于对该存储的图像进行识别，得到该图像包含指定内容的估分值的识别结果。
该第二回调策略包括将估分值在第三预设分值范围内的图像封禁或删除。
该图像识别处理模块504还用于判断该存储的图像对应的估分值是否在第三预设分值范围内，以及若该存储的图像的估分值在该第三预设分值范围内，则将该存储的图像封禁或删除，得到封禁结果或删除结果。
该回调通知模块508还用于将该封禁结果或删除结果回调通知该请求方。
封禁结果可为标记为违法图像且无法显示。删除结果为通知图像已删除。
在一个实施例中，当获取请求方发送包含某图像的图像识别请求在预设时间内未返回识别结果时，图像识别接口模块502获取请求方再次发送包含所述图像的图像识别请求，且再次获取包含所述图像的图像识别请求不超过预设重试次数。预设重试次数可根据情况设定，如3次或5次等。
下面结合具体的应用场景描述图像识别回调通知的方法和装置的实现过程。以图像识别回调通知的方法应用于色情图像识别为例，色情图像识别的架构如图7所示，包括回调策略配置模块702、色情图像识别处理模块704、色情图像识别接口706、回调通知模块708和存储模块710。回调策略配置模块702获取用户在回调策略配置界面输入的数据，并根据该输入的数据生成回调策略。色情图像识别接口704接收请求方发送的色情图像识别请求，并将该色情图像识别请求中的图像发送给色情图像识别处理模块704。色情图像识别处理模块704对该图像进行识别，得出该图像的色情分值，再读取请求方配置的回调策略，根据回调策略判断图像的色情分值是否在预设分值范围内，若是，则将图像的识别结果发送至回调通知模块708。回调通知模块708用于将图像的识别结果回调通知给请求方或者请求方配置的服务器。
存储模块710用于存储请求方通过第三方应用上传的图像。色情图像识别处理模块704对存储模块710中存储的图像进行识别，得到图像的色情分 值，获取请求方的回调策略，根据回调策略对图像进行封禁或删除等操作，得到对应的封禁结果或删除结果，将封禁结果或删除结果发送至回调通知模块708，并将需要回调的图像信息发送给回调通知模块708。回调通知模块708将图像的统一资源定位符、色情分值、是否被封禁或删除等识别结果，通知给请求方或请求方配置的服务器。若请求方发送的图像识别请求失败，会进行多次重试。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程，是可以通过计算机程序来指令相关的硬件来完成，所述的程序可存储于一非易失性计算机可读取存储介质中，该程序在执行时，可包括如上述各方法的实施例的流程。其中，所述的存储介质可为磁碟、光盘、只读存储记忆体(Read-Only Memory，ROM)等。
以上所述实施例仅表达了本发明的几种实施方式，其描述较为具体和详细，但并不能因此而理解为对本发明专利范围的限制。应当指出的是，对于本领域的普通技术人员来说，在不脱离本发明构思的前提下，还可以做出若干变形和改进，这些都属于本发明的保护范围。因此，本发明专利的保护范围应以所附权利要求为准。

Claims (21)

  1. 一种图像识别回调通知的方法，包括：
    获取请求方发送的请求方标识及对应的图像识别请求；
    根据所述图像识别请求对图像进行识别，得到识别结果；
    获取与所述请求方标识对应的第一回调策略；及
    当根据所述第一回调策略确定将所述识别结果进行回调时，将所述图像的识别结果回调通知给所述请求方。
  2. 根据权利要求1所述的方法，其特征在于，所述根据所述图像识别请求对图像进行识别，得到识别结果包括：
    根据所述图像识别请求对图像进行识别，得到所述图像包含指定内容的估分值的识别结果；
    所述第一回调策略为估分值在第一预设分值范围内时进行回调通知；
    所述当根据所述第一回调策略确定将所述识别结果进行回调时，将所述图像的识别结果回调通知给所述请求方包括：
    当所述图像对应的估分值在所述第一预设分值范围内，则将所述图像的识别结果回调通知给所述请求方。
  3. 根据权利要求1所述的方法，其特征在于，所述根据所述图像识别请求对图像进行识别，得到识别结果包括：
    根据所述图像识别请求对图像进行识别，得到所述图像包含指定内容的估分值的识别结果；
    所述第一回调策略为在预设时间范围内估分值从高到低中预设数量的图像的识别结果进行回调；
    所述当根据所述第一回调策略确定将所述识别结果进行回调时，将所述图像的识别结果回调通知给所述请求方包括：
    根据所述第一回调策略获取在预设时间范围内估分值从高到低中预设数量的图像的识别结果；
    将所述估分值从高到低中预设数量的图像的识别结果回调通知给所述请求方。
  4. 根据权利要求1所述的方法，其特征在于，还包括：
    接收请求方上传的请求方标识及对应的图像，并存储所述请求方标识及对应的图像。
  5. 根据权利要求4所述的方法，其特征在于，还包括：
    对所述存储的图像进行识别，得到识别结果；
    获取与所述请求方标识对应的第二回调策略；
    当根据所述第二回调策略确定将所述存储的图像的识别结果进行回调时，将所述存储的图像的识别结果回调通知给所述请求方。
  6. 根据权利要求5所述的方法，其特征在于，所述对所述存储的图像进行识别，得到识别结果包括：
    对所述存储的图像进行识别，得到所述图像包含指定内容的估分值的识别结果；
    所述第二回调策略包括估分值在第二预设分值范围内时进行回调通知；
    所述当根据所述第二回调策略确定将所述存储的图像的识别结果进行回调时，将所述存储的图像的识别结果回调通知给所述请求方，包括：
    当所述存储的图像的估分值在所述第二预设分值范围内，将所述存储的图像的识别结果回调通知给所述请求方。
  7. 根据权利要求5所述的方法，其特征在于，所述对所述存储的图像进行识别，得到识别结果包括：
    对所述存储的图像进行识别，得到所述图像包含指定内容的估分值的识别结果；
    所述第二回调策略包括将估分值在第三预设分值范围内的图像封禁或删除；
    所述方法还包括：
    当所述存储的图像的估分值在所述第三预设分值范围内，将所述存储的图像封禁或删除，得到封禁结果或删除结果，并将所述封禁结果或删除结果回调通知所述请求方。
  8. 根据权利要求5所述的方法，其特征在于，所述对所述存储的图像进行识别，得到识别结果包括：
    对所述存储的图像进行识别，得到所述图像包含指定内容的估分值的识别结果；及
    所述第二回调策略包括在预设时间范围内估分值从高到低中预设数量的图像的识别结果进行回调；
    所述当根据所述第二回调策略确定将所述存储的图像的识别结果进行回调时，将所述存储的图像的识别结果回调通知给所述请求方包括：
    根据所述第二回调策略获取在预设时间范围内估分值从高到低中预设数量的图像的识别结果；
    将所述估分值从高到低中预设数量的图像的识别结果回调通知给所述请求方。
  9. 根据权利要求5所述的方法，其特征在于，所述对所述存储的图像进行识别，得到识别结果包括：
    对所述存储的图像进行识别，得到所述图像包含指定内容的估分值的识别结果；
    所述第二回调策略包括估分值在第二预设分值范围内时进行回调通知且将估分值在第三预设分值范围内的图像封禁或删除；
    所述当根据所述第二回调策略确定将所述存储的图像的识别结果进行回调时，将所述存储的图像的识别结果回调通知给所述请求方包括：
    当所述存储的图像的估分值在所述第二预设分值范围内，则将所述存储的图像的识别结果回调通知给所述请求方；
    当所述存储的图像的估分值在所述第三预设分值范围内，则将所述存储的图像封禁或删除，得到封禁结果或删除结果，并将所述封禁结果或删除结 果回调通知所述请求方。
  10. 根据权利要求1所述的方法，其特征在于，还包括：
    当获取请求方发送包含某图像的图像识别请求在预设时间内未返回识别结果时，获取请求方再次发送包含所述图像的图像识别请求，且再次获取包含所述图像的图像识别请求不超过预设重试次数。
  11. 一种服务器，包括存储器及处理器，所述存储器中储存有计算机可读指令，所述指令被所述处理器执行时，使得所述处理器执行以下步骤：
    获取请求方发送的请求方标识及对应的图像识别请求；
    根据所述图像识别请求对图像进行识别，得到识别结果；
    获取与所述请求方标识对应的第一回调策略；及
    根据所述第一回调策略确定将所述识别结果进行回调时，将所述图像的识别结果回调通知给所述请求方。
  12. 根据权利要求11所述的服务器，其特征在于，所述处理器所执行的所述根据所述图像识别请求对图像进行识别，得到识别结果包括：
    根据所述图像识别请求对图像进行识别，得到所述图像包含指定内容的估分值的识别结果；
    所述第一回调策略为估分值在第一预设分值范围内时进行回调通知；
    所述根据所述第一回调策略确定将所述识别结果进行回调时，将所述图像的识别结果回调通知给所述请求方包括：
    当所述图像对应的估分值在所述第一预设分值范围内，则将所述图像的识别结果回调通知给所述请求方。
  13. 根据权利要求11所述的服务器，其特征在于，所述处理器所执行的所述根据所述图像识别请求对图像进行识别，得到识别结果包括：
    根据所述图像识别请求对图像进行识别，得到所述图像包含指定内容的估分值的识别结果；
    所述第一回调策略为在预设时间范围内估分值从高到低中预设数量的图像的识别结果进行回调；
    所述根据所述第一回调策略确定将所述识别结果进行回调时，将所述图像的识别结果回调通知给所述请求方包括：
    根据所述第一回调策略获取在预设时间范围内估分值从高到低中预设数量的图像的识别结果；
    将所述估分值从高到低中预设数量的图像的识别结果回调通知给所述请求方。
  14. 根据权利要求11所述的服务器，其特征在于，所述处理器还用于执行以下步骤：
    接收请求方上传的请求方标识及对应的图像，并存储所述请求方标识及对应的图像。
  15. 根据权利要求14所述的服务器，其特征在于，所述处理器还用于执行以下步骤：
    对所述存储的图像进行识别，得到识别结果；
    获取与所述请求方标识对应的第二回调策略；
    当根据所述第二回调策略确定将所述存储的图像的识别结果进行回调时，将所述存储的图像的识别结果回调通知给所述请求方。
  16. 根据权利要求15所述的服务器，其特征在于，所述处理器所执行的所述对所述存储的图像进行识别，得到识别结果包括：
    对所述存储的图像进行识别，得到所述图像包含指定内容的估分值的识别结果；
    所述第二回调策略包括估分值在第二预设分值范围内时进行回调通知；
    所述处理器所执行的所述当根据所述第二回调策略确定将所述存储的图像的识别结果进行回调时，将所述存储的图像的识别结果回调通知给所述请求方，包括：
    当所述存储的图像的估分值在所述第二预设分值范围内，则将所述存储的图像的识别结果回调通知给所述请求方。
  17. 根据权利要求15所述的服务器，其特征在于，所述处理器所执行的 所述对所述存储的图像进行识别，得到识别结果包括：
    对所述存储的图像进行识别，得到所述图像包含指定内容的估分值的识别结果；
    所述第二回调策略包括将估分值在第三预设分值范围内的图像封禁或删除；
    所述处理器还用于执行以下步骤：
    当所述存储的图像的估分值在所述第三预设分值范围内，则将所述存储的图像封禁或删除，得到封禁结果或删除结果，并将所述封禁结果或删除结果回调通知所述请求方。
  18. 根据权利要求15所述的服务器，其特征在于，所述处理器所执行的所述对所述存储的图像进行识别，得到识别结果包括：
    对所述存储的图像进行识别，得到所述图像包含指定内容的估分值的识别结果；及
    所述第二回调策略包括在预设时间范围内估分值从高到低中预设数量的图像的识别结果进行回调；
    所述处理器所执行的所述当根据所述第二回调策略确定将所述存储的图像的识别结果进行回调时，将所述存储的图像的识别结果回调通知给所述请求方包括：
    根据所述第二回调策略获取在预设时间范围内估分值从高到低中预设数量的图像的识别结果；
    将所述估分值从高到低中预设数量的图像的识别结果回调通知给所述请求方。
  19. 根据权利要求15所述的服务器，其特征在于，所述处理器所执行的所述对所述存储的图像进行识别，得到识别结果包括：
    对所述存储的图像进行识别，得到所述图像包含指定内容的估分值的识别结果；
    所述第二回调策略包括估分值在第二预设分值范围内时进行回调通知且 将估分值在第三预设分值范围内的图像封禁或删除；
    所述处理器所执行的所述当根据所述第二回调策略确定将所述存储的图像的识别结果进行回调时，将所述存储的图像的识别结果回调通知给所述请求方包括：
    当所述存储的图像的估分值在所述第二预设分值范围内，则将所述存储的图像的识别结果回调通知给所述请求方；
    当所述存储的图像的估分值在所述第三预设分值范围内，则将所述存储的图像封禁或删除，得到封禁结果或删除结果，并将所述封禁结果或删除结果回调通知所述请求方。
  20. 根据权利要求11所述的服务器，其特征在于，所述处理器还用于执行以下步骤：
    当获取请求方发送包含某图像的图像识别请求在预设时间内未返回识别结果时，获取请求方再次发送包含所述图像的图像识别请求，且再次获取包含所述图像的图像识别请求不超过预设重试次数。
  21. 一个或多个存储有计算机可执行指令的非易失性可读存储介质，所述计算机可执行指令被一个或多个处理器执行时，使得所述一个或多个处理器执行以下步骤如权利要求1至10任一项所述的图像识别回调通知的方法。
PCT/CN2017/092557 2016-07-12 2017-07-12 图像识别回调通知的方法、服务器和计算机可读存储介质 WO2018010652A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201610548024.X 2016-07-12
CN201610548024.XA CN106203514B (zh) 2016-07-12 2016-07-12 图像识别回调通知的方法和装置

Publications (1)

Publication Number Publication Date
WO2018010652A1 true WO2018010652A1 (zh) 2018-01-18

Family

ID=57476565

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/092557 WO2018010652A1 (zh) 2016-07-12 2017-07-12 图像识别回调通知的方法、服务器和计算机可读存储介质

Country Status (2)

Country Link
CN (1) CN106203514B (zh)
WO (1) WO2018010652A1 (zh)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106203514B (zh) * 2016-07-12 2019-02-12 腾讯科技（深圳）有限公司 图像识别回调通知的方法和装置
CN110399573A (zh) * 2018-04-16 2019-11-01 中国移动通信有限公司研究院 一种信息处理方法、装置、设备及计算机可读存储介质

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102682091A (zh) * 2012-04-25 2012-09-19 腾讯科技（深圳）有限公司 基于云服务的视觉搜索方法和系统
CN103051705A (zh) * 2012-12-19 2013-04-17 中兴通讯股份有限公司 目标人的确定方法、装置及移动终端
CN105095848A (zh) * 2014-12-01 2015-11-25 北京云识图信息技术有限公司 目标识别的方法及系统
CN105404631A (zh) * 2014-09-15 2016-03-16 腾讯科技（深圳）有限公司 图片识别方法和装置
WO2016063484A1 (ja) * 2014-10-23 2016-04-28 日本電気株式会社 画像処理装置、表示制御装置、画像処理方法、および、記録媒体
CN106203514A (zh) * 2016-07-12 2016-12-07 腾讯科技（深圳）有限公司 图像识别回调通知的方法和装置

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110047388A1 (en) * 2009-08-24 2011-02-24 Samsung Electronics Co., Ltd. Method and apparatus for remotely controlling access to pornographic content of an image
JP2012079210A (ja) * 2010-10-05 2012-04-19 Hitachi Ltd サービス連携システム
CN104281833B (zh) * 2013-07-08 2018-12-18 深圳市腾讯计算机系统有限公司 色情图像识别方法和装置
CN104935648B (zh) * 2015-06-03 2018-07-17 北京快网科技有限公司 一种高性价比的cdn系统及文件预推、分片缓存的方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102682091A (zh) * 2012-04-25 2012-09-19 腾讯科技（深圳）有限公司 基于云服务的视觉搜索方法和系统
CN103051705A (zh) * 2012-12-19 2013-04-17 中兴通讯股份有限公司 目标人的确定方法、装置及移动终端
CN105404631A (zh) * 2014-09-15 2016-03-16 腾讯科技（深圳）有限公司 图片识别方法和装置
WO2016063484A1 (ja) * 2014-10-23 2016-04-28 日本電気株式会社 画像処理装置、表示制御装置、画像処理方法、および、記録媒体
CN105095848A (zh) * 2014-12-01 2015-11-25 北京云识图信息技术有限公司 目标识别的方法及系统
CN106203514A (zh) * 2016-07-12 2016-12-07 腾讯科技（深圳）有限公司 图像识别回调通知的方法和装置

Also Published As

Publication number Publication date
CN106203514A (zh) 2016-12-07
CN106203514B (zh) 2019-02-12

Similar Documents

Publication Publication Date Title
JP6417154B2 (ja) 情報処理装置、情報処理方法、及びプログラム
CN109618173B (zh) 视频压缩方法、装置和计算机可读存储介质
US9177194B2 (en) System and method for visually distinguishing faces in a digital image
US8983152B2 (en) Image masks for face-related selection and processing in images
WO2018036462A1 (zh) 图像分割的方法、计算机设备及存储介质
US7827412B2 (en) Content presenting method
US20160219010A1 (en) Claiming conversations between users and non-users of a social networking system
US20140198982A1 (en) System and method for recognizing offensive images
CN107220614B (zh) 图像识别方法、装置及计算机可读存储介质
WO2018010652A1 (zh) 图像识别回调通知的方法、服务器和计算机可读存储介质
WO2022111269A1 (zh) 视频的细节增强方法、装置、移动终端和存储介质
US8265383B2 (en) Method and communication terminal for searching phone book for desired contacts
CN113923474B (zh) 视频帧处理方法、装置、电子设备以及存储介质
CN114040048A (zh) 一种隐私保护方法及电子设备
CN103353879B (zh) 图像处理方法及设备
US20170171462A1 (en) Image Collection Method, Information Push Method and Electronic Device, and Mobile Phone
CN109147911A (zh) 病症信息显示方法及装置、计算机可读存储介质
CN112163200A (zh) 图片处理方法、装置及电子设备
US20180012066A1 (en) Photograph processing method and system
CN112215237B (zh) 图像处理方法及装置、电子设备和计算机可读存储介质
CN114303352B (zh) 推送内容的处理方法、装置、电子设备以及存储介质
CN113869002A (zh) 一种字符显示方法、装置、电子设备及存储介质
CN113487497A (zh) 图像处理方法、装置和电子设备
CN112929489A (zh) 通信方法及装置
CN113440843B (zh) 云游戏启动控制方法、装置、云服务器和终端设备

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: 17826988

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: 17826988

Country of ref document: EP

Kind code of ref document: A1