WO2019085589A1 - 无人值守场景中非法行为的识别方法和装置 - Google Patents

无人值守场景中非法行为的识别方法和装置 Download PDF

Info

Publication number
WO2019085589A1
WO2019085589A1 PCT/CN2018/100977 CN2018100977W WO2019085589A1 WO 2019085589 A1 WO2019085589 A1 WO 2019085589A1 CN 2018100977 W CN2018100977 W CN 2018100977W WO 2019085589 A1 WO2019085589 A1 WO 2019085589A1
Authority
WO
WIPO (PCT)
Prior art keywords
user
data
behavior
mouth
limb
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/CN2018/100977
Other languages
English (en)
French (fr)
Inventor
贺武杰
曾晓东
张晓博
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.)
Alibaba Group Holding Ltd
Original Assignee
Alibaba Group Holding 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 Alibaba Group Holding Ltd filed Critical Alibaba Group Holding Ltd
Priority to KR1020207010861A priority Critical patent/KR102256895B1/ko
Priority to KR1020217014172A priority patent/KR102429739B1/ko
Priority to SG11202003011TA priority patent/SG11202003011TA/en
Publication of WO2019085589A1 publication Critical patent/WO2019085589A1/zh
Priority to US16/796,616 priority patent/US10783362B2/en
Anticipated expiration legal-status Critical
Priority to US16/943,639 priority patent/US10990813B2/en
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/168—Feature extraction; Face representation
    • G06V40/171—Local features and components; Facial parts ; Occluding parts, e.g. glasses; Geometrical relationships
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; 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 OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/50—Context or environment of the image
    • G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/161—Detection; Localisation; Normalisation
    • G06V40/165—Detection; Localisation; Normalisation using facial parts and geometric relationships
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/161—Detection; Localisation; Normalisation
    • G06V40/167—Detection; Localisation; Normalisation using comparisons between temporally consecutive images
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/20—Movements or behaviour, e.g. gesture recognition
    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING SYSTEMS, e.g. PERSONAL CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B13/00—Burglar, theft or intruder alarms
    • G08B13/02—Mechanical actuation
    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING SYSTEMS, e.g. PERSONAL CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B13/00—Burglar, theft or intruder alarms
    • G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
    • G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
    • G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
    • G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
    • G08B13/19602—Image analysis to detect motion of the intruder, e.g. by frame subtraction
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06K—GRAPHICAL DATA READING; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
    • G06K19/00—Record carriers for use with machines and with at least a part designed to carry digital markings
    • G06K19/06—Record carriers for use with machines and with at least a part designed to carry digital markings characterised by the kind of the digital marking, e.g. shape, nature, code
    • G06K19/067—Record carriers with conductive marks, printed circuits or semiconductor circuit elements, e.g. credit or identity cards also with resonating or responding marks without active components
    • G06K19/07—Record carriers with conductive marks, printed circuits or semiconductor circuit elements, e.g. credit or identity cards also with resonating or responding marks without active components with integrated circuit chips
    • G06K19/0723—Record carriers with conductive marks, printed circuits or semiconductor circuit elements, e.g. credit or identity cards also with resonating or responding marks without active components with integrated circuit chips the record carrier comprising an arrangement for non-contact communication, e.g. wireless communication circuits on transponder cards, non-contact smart cards or RFIDs
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V2201/00—Indexing scheme relating to image or video recognition or understanding
    • G06V2201/07—Target detection

Definitions

  • the present specification relates to the field of Internet technologies, and in particular, to a method and apparatus for identifying illegal behavior in an unattended scenario.
  • unattended scenes are increasingly applied to daily life, such as unmanned supermarkets, unmanned gyms, and unmanned KTV.
  • unattended scenes are also facing some illegal behaviors, such as: eating and drinking in an unmanned supermarket, not paying to leave, destroying the fitness facilities in the unmanned gym. Therefore, there is a need to provide an effective illegal behavior recognition scheme in an unattended scenario.
  • the present specification provides a method and apparatus for identifying illegal behavior in an unattended scenario.
  • a method for identifying illegal behavior in an unattended scenario comprising:
  • An apparatus for identifying illegal behavior in an unattended scenario comprising:
  • the first collecting unit collects the limb data of the user located in the unattended scene
  • a second collecting unit collecting feature data of the object in the unattended scene
  • the behavior recognition unit identifies whether the user has an illegal behavior based on the limb data and the feature data of the object.
  • An apparatus for identifying illegal behavior in an unattended scenario comprising:
  • a memory for storing machine executable instructions
  • the processor is caused to:
  • the present specification can collect the limb data of the user and the feature data of the object in the unattended scene, and integrate the actual situation of the user and the object to identify the illegal behavior, thereby realizing the illegal behavior in the unattended scene. Effective identification.
  • FIG. 1 is a schematic flow chart of a method for identifying illegal behavior in an unattended scenario, according to an exemplary embodiment of the present specification.
  • FIG. 2 is a schematic flow chart of a method for identifying illegal behavior in another unattended scenario according to an exemplary embodiment of the present specification.
  • FIG. 3 is a schematic flow chart of a method for identifying illegal behavior in another unattended scenario according to an exemplary embodiment of the present specification.
  • FIG. 4 is a schematic structural diagram of an apparatus for identifying illegal behavior in an unattended scenario, according to an exemplary embodiment of the present specification.
  • FIG. 5 is a block diagram of an apparatus for identifying illegal behavior in an unattended scenario, according to an exemplary embodiment of the present specification.
  • first, second, third, etc. may be used in this specification to describe various information, such information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other.
  • first information may also be referred to as the second information without departing from the scope of the present description.
  • second information may also be referred to as the first information.
  • word "if” as used herein may be interpreted as "when” or “when” or “in response to a determination.”
  • FIG. 1 is a schematic flow chart of a method for identifying illegal behavior in an unattended scenario, according to an exemplary embodiment of the present specification.
  • the method for identifying an illegal behavior in the unattended scenario may be applied to a front-end identification device in an unattended scenario, such as an identification device having a processing function deployed in an unattended scenario; the identification method may also be applied to none
  • the back-end identification device in the scene, such as the background recognition device, is not limited in this specification.
  • the method for identifying illegal behavior in an unattended scenario may include the following steps:
  • step 102 the limb data of the user located in the unattended scene is collected.
  • the collection of the limb data can be started after the user enters the unattended scene.
  • the limb data may include: the user's limb behavior data and the limb position data, and the user's limbs may include: limbs, mouths, and the like.
  • the limb data of the user in the unattended scene can be collected by the machine vision technology.
  • the limb data can also be collected by other techniques, which is not specifically limited in this specification.
  • Step 104 Collect feature data of the object in the unattended scene.
  • the collection of the feature data may also be started after the user enters the unattended scenario.
  • the object is usually an object in an unattended scene, for example, goods in an unmanned supermarket, containers in which an unmanned supermarket is placed, and fitness equipment in an unmanned gym.
  • the feature data may include: position data of the object, an outer surface image of the object, oscillating data collected by a sensor disposed in the object, and the like.
  • the feature data of the object in the unattended scene can also be collected by the machine vision technology.
  • the feature data can also be collected by other means such as a sensor, and the specification does not particularly limit this.
  • Step 106 Identify whether the user has an illegal behavior according to the limb data and the feature data of the object.
  • the collected limb data and the feature data may be analyzed to determine whether the user in the unattended scene has an illegal behavior, for example, destroying an item or the like.
  • the present specification can collect the limb data of the user and the feature data of the object in the unattended scene, and integrate the actual situation of the user and the object to identify the illegal behavior, thereby realizing the illegal behavior in the unattended scene. Effective identification.
  • the following is an unmanned scene, which is an unmanned supermarket.
  • the illegal implementation of illegal eating, drinking, and destroying containers is used as an example to describe the implementation process of this manual.
  • the user may start collecting the limb data and the object feature data after the user enters the unmanned supermarket, to determine whether the user has an illegal behavior, and may stop after the user leaves the unmanned supermarket. Acquisition of the limb data and feature data.
  • the method for judging the user entering and leaving the unmanned supermarket is various.
  • the user may be determined to enter the unmanned supermarket, and the user may be determined to leave the unmanned supermarket after the user internally triggers the opening of the gate of the unmanned supermarket.
  • the user may enter the unmanned supermarket after the identification device in the unmanned supermarket identifies the living person, and may determine that the user leaves the unmanned supermarket after the identification device in the unmanned supermarket cannot recognize the living person.
  • an unmanned supermarket when a user enters an unmanned supermarket, they may eat and drink in the supermarket, leave after eating and drinking, and bring economic losses to the unmanned supermarket.
  • the identification method provided in this embodiment may include the following steps:
  • Step 202 Collect mouth position data of the user located in the unmanned supermarket.
  • the face feature of the user may be collected first, and then the mouth of the user is recognized, thereby obtaining the mouth position data of the user.
  • the position data of the user's mouth can be calculated based on the relative positional relationship between the user's mouth and the collection device and the position data of the collection device.
  • the mouth position data of the user may be collected in other manners, which is not specifically limited in this embodiment.
  • the mouth position data may be latitude and longitude data of the mouth.
  • the mouth position data may also be relative position data of the mouth in the unmanned supermarket.
  • a spatial rectangular coordinate system may be established with a certain spatial point in the unmanned supermarket (for example, a wall foot of an unmanned supermarket), and then the coordinates of the user's mouth in the rectangular coordinate system of the space are used as the mouth position data. This specification does not specifically limit this.
  • Step 204 Collect object position data of the object in the unmanned supermarket.
  • the object when identifying illegal eating and drinking, the object is usually a product sold in an unmanned supermarket.
  • the position data of the product can be determined by the RFID positioning technology. If the RFID tag is not provided on the product, the location data of the product can be determined by other positioning technologies, which is not specifically limited in this embodiment.
  • RFID Radio Frequency Identification
  • the initial position of each item can be stored in advance, for example, the position of the item on the shelf can be used as the initial position.
  • the subsequent steps of this embodiment can be performed when the location of a certain item is not in its initial position.
  • the object position data may be latitude and longitude data, or may be the relative position data of the object in the unmanned supermarket, which is not particularly limited in this embodiment.
  • Step 206 Calculate a distance between the mouth of the user and the object according to the mouth position data and the object position data.
  • the mouth position data and the object position data collected at the same time point are generally used to ensure the accuracy of the above distance.
  • Step 208 When the distance between the mouth and the object is within a distance threshold, the user is determined to have an illegal eating and drinking behavior.
  • the distance threshold may be preset, for example, 20 cm, 30 cm, or the like.
  • the user's mouth When the user's mouth is within the above distance threshold from the goods, it usually indicates that the user's mouth is close to the goods. The user may smell the goods, and the user may steal or steal the goods.
  • the distance between the user's mouth and the goods within the above distance threshold is usually short; if the user is stealing or stealing the goods, the distance between the user's mouth and the goods is as described above.
  • the duration of the distance threshold is usually long. Therefore, when the length of the user's mouth and the product within the distance threshold reaches a predetermined length of time, it is determined that the user is stealing or stealing, that is, the user has an illegal eating and drinking behavior.
  • the predetermined duration may be set in advance, for example, 5 seconds, 10 seconds, or the like.
  • the unattended scenario when the distance between the user's mouth and the product within the distance threshold reaches a predetermined length of time in the unattended scenario, it is determined that the user has an illegal eating and drinking behavior, thereby implementing the unattended scene. Identification of illegal eating and drinking behavior.
  • the user's mouth behavior data may also be collected as the limb data, and it is determined whether the mouth behavior data matches the predetermined mouth movement law. For example: chewing rules, swallowing rules, etc.
  • the collected mouth behavior data matches the predetermined mouth movement law, and the distance between the mouth of the user and the goods is within a predetermined time period, it may be determined that the user is stealing or stealing, that is, the user has illegal eating and drinking. behavior.
  • the collected mouth behavior data does not match the predetermined mouth movement law, it can be determined that the user does not have an illegal eating and drinking behavior.
  • the determination of whether the user's mouth behavior data matches the predetermined mouth motion law may be performed before step 206 shown in FIG. 2, or after step 206, and may also be performed simultaneously with step 206. This is not a special limitation.
  • the attribute category data of the goods may also be collected as the characteristic data of the goods.
  • the attribute category data usually refers to the category classification of the goods, for example: edible, inedible, and the like.
  • the product selected by the user must be an edible category. Therefore, it is determined that the user is stealing or stealing when the length of the user's mouth and the article within the distance threshold reaches a predetermined length of time and the attribute category of the item belongs to the edible category. For example, if the attribute category of the product is bottled drinking water, and the user's distance from the bottled drinking water is within a predetermined time period from the threshold, it can be determined that the user is stealing the bottled drinking water.
  • the attribute category of the item is a household item, and although the distance between the user and the household item within the distance threshold reaches a predetermined length of time, it may be determined that the user does not have an illegal eating and drinking behavior.
  • the determination of the attribute category of the product may be performed before the step 206 shown in FIG. 2, or after the step 206, and may be performed simultaneously with the step 206. This embodiment does not particularly limit this.
  • the outer surface image data of the goods may also be collected as the feature data of the goods, and the outer surface image data can generally be understood as the outer package image of the goods.
  • the image data of the outer surface of the product matches its default outer surface image.
  • the default outer surface image can be pre-stored. If the image data of the outer surface of the product matches the default outer surface image, it can be said that the outer packaging of the product is not damaged; if the image data of the outer surface of the product does not match the default outer surface image, it can be explained that the outer surface of the product is outside The packaging has been destroyed.
  • the matching of the external surface image data of the product may be performed before the step 206 shown in FIG. 2, or after the step 206, and may be performed simultaneously with the step 206. This embodiment does not particularly limit this.
  • the goods of the unmanned supermarket may be destroyed, for example, by kicking the container, etc., causing economic losses to the unmanned supermarket.
  • the identification method provided in this embodiment may include the following steps:
  • step 302 the limb behavior data of the user located in the unmanned supermarket is collected.
  • the limb behavior data may include: arm behavior data, leg behavior data, and the like.
  • the behavior data of the limbs of the user over a period of time may be collected for analyzing the movement of the limbs of the user.
  • step 304 the oscillating data of the unmanned supermarket container is collected.
  • sensors such as gyroscope sensors, acceleration sensors, etc.
  • sensors may be disposed on the container of the supermarket to collect the shock data of the container.
  • the shock data of the container can also be collected by other technologies such as machine vision.
  • Step 306 When the behavior data of the limb of the user matches a predetermined motion law of the limbs, and the container is oscillated, it is determined that the user has an illegal behavior.
  • step 302 after collecting the behavior data of the limbs of the user, it can be determined whether the behavior data of the limbs matches the predetermined motion law of the limbs, for example, a “ ⁇ ” rule, a “kick” rule, a “ ⁇ ” rule, and the like.
  • the behavior data of the limbs matches the predetermined motion law of the limbs, it indicates that the user may destroy the items in the unmanned supermarket, and the judgment may continue.
  • the behavior data of the limbs does not match the predetermined motion law of the limbs, it may be stated that the user does not have the illegal behavior of destroying the items in the unmanned supermarket.
  • step 304 after collecting the oscillating data of the container, whether the container is oscillated according to the oscillating data may be determined.
  • the container is oscillated and the user's limb behavior data matches the predetermined limb movement pattern, it can be determined that the user has an illegal act of destroying the container.
  • the container does not oscillate, then it can be determined that the user does not have an illegal act of destroying the container.
  • the user's limb behavior data matches the predetermined law of the extremity of the extremities, and when the object is oscillated, the user is determined to have an illegal behavior of destroying the object, thereby achieving Identification of illegal vandalism in a person's duty scene.
  • the distance between the limb that the user matches the motion law of the limbs and the object may be calculated, and if the distance is within a predetermined range, that is, the distance is In the near future, it is possible to determine the behavior of the user who illegally destroys the object.
  • the user may also damage the goods sold by the unmanned supermarket, such as tearing off the outer packaging of the goods. Therefore, when the identification of the illegal destruction behavior is performed, the outer surface image data of the object may also be collected, and then the collected outer surface image data and the default outer surface image of the object are matched to determine whether the outer packaging of the object is damage.
  • the image data of the outer surface can also be matched with the default outer surface image, which is not particularly limited in this embodiment.
  • the external surface image may not necessarily change. In order to improve the recognition accuracy, it is usually necessary to collect other data for identification.
  • the developer in implementing the above various identification schemes, it is usually required that the developer first establish a corresponding behavior recognition model, and then improve the behavior recognition model by continuously training the model. Before the behavior recognition model is perfected, if it is impossible to determine whether the user has an illegal behavior according to the user's limb data and the feature data of the object, the judgment request may be output to the relevant staff member, and the staff member manually judges and judges the result. Return to refine the above behavioral recognition model.
  • the above-mentioned limb data and feature data can be continuously collected, and the recognition is continued, and if the certain length of time is reached, the recognition is still impossible. It is also possible to output a judgment request to the relevant staff member to request manual intervention.
  • the user's illegal behavior can be alerted and disciplined in a variety of different ways.
  • the user may be subject to credit discipline when the user has an illegal behavior.
  • the user before entering the unattended scenario, the user usually has to be authenticated.
  • the QR code is scanned for the unattended supermarket to obtain the account information of the user.
  • the credit information of the user account of the user may be updated negatively, for example, the credit score thereof is lowered.
  • the illegal behavior can be broadcast in the unattended scene.
  • the illegal behavior of the user may be broadcasted by audio, or the illegal behavior of the user may be played back through video for warning.
  • the amount of resources corresponding to the object for which the illegal behavior is directed may also be added to the user's bill.
  • the amount of resources generally refers to the value of the object.
  • the amount of potato chips can be added to the user's bill, even if the user drops the package in the unmanned supermarket after eating the potato chips, The user charges the corresponding fee.
  • the present specification also provides an embodiment of an apparatus for identifying illegal behavior in an unattended scenario.
  • Embodiments of the identification device for illegal behavior in an unattended scenario of the present specification can be applied to an identification device.
  • the device embodiment may be implemented by software, or may be implemented by hardware or a combination of hardware and software.
  • the processor of the identification device is formed by reading the corresponding computer program instructions in the non-volatile memory into the memory.
  • FIG. 4 a hardware structure diagram of the identification device of the identification device for illegal behavior in the unattended scenario of the present specification, except for the processor, the memory, the network interface, and the processor shown in FIG.
  • the identification device in which the device is located in the embodiment may generally include other hardware according to the actual function of the identification device, and details are not described herein.
  • FIG. 5 is a block diagram of an apparatus for identifying illegal behavior in an unattended scenario, according to an exemplary embodiment of the present specification.
  • the identification device 400 for the illegal behavior in the unattended scenario may be applied to the identification device shown in FIG. 4, including: a first collection unit 401, a second collection unit 402, and a behavior recognition unit 403.
  • the first collecting unit 401 collects limb data of the user in the unattended scene
  • the second collecting unit 402 collects feature data of the object in the unattended scene
  • the behavior recognition unit 403 identifies whether the user has an illegal behavior based on the limb data and the feature data of the object.
  • the limb data includes: the mouth position data of the user;
  • the feature data includes: object position data of the object;
  • the behavior recognition unit 403 calculates a distance between the mouth of the user and the object according to the mouth position data and the object position data; when the distance between the mouth and the object is within a distance threshold When the time is predetermined, it is determined that the user has an illegal behavior.
  • the limb data further includes: the mouth behavior data of the user;
  • the behavior recognition unit 403 determines that the user is illegal when the length of the mouth and the object is within a distance threshold for a predetermined length of time, and the mouth behavior data of the user matches a predetermined movement rule of the mouth. behavior.
  • the feature data further includes: attribute category data of the object;
  • the behavior recognition unit 403 determines that the user has an illegal behavior when the length of the mouth from the object is within a distance threshold for a predetermined length of time and the attribute category of the object belongs to an edible category.
  • the feature data further includes: image data of an outer surface of the object;
  • the behavior recognition unit 403 determines that the user is illegal when the distance between the mouth and the object is within a distance threshold for a predetermined length of time, and the outer surface image of the object does not match the default outer surface image. behavior.
  • the limb data includes: behavior data of the limbs of the user;
  • the feature data includes: oscillating data of the object
  • the behavior recognition unit 403 determines that the user has an illegal behavior when the limb behavior data of the user matches a predetermined motion law of the limbs and the object is oscillated.
  • the oscillating data is collected by a sensor disposed on the object.
  • the credit disciplinary unit 404 when the user has an illegal behavior, performs credit punishment on the user.
  • the broadcast disciplinary unit 405 broadcasts the illegal behavior in the unattended scene when the user has an illegal behavior.
  • the resource discerning unit 406 adds the resource amount corresponding to the object to the user's bill when the user has an illegal behavior.
  • the device embodiment since it basically corresponds to the method embodiment, reference may be made to the partial description of the method embodiment.
  • the device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, ie may be located A place, or it can be distributed to multiple network units. Some or all of the modules may be selected according to actual needs to achieve the objectives of the present specification. Those of ordinary skill in the art can understand and implement without any creative effort.
  • the system, device, module or unit illustrated in the above embodiments may be implemented by a computer chip or an entity, or by a product having a certain function.
  • a typical implementation device is a computer, and the specific form of the computer may be a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email transceiver, and a game control.
  • the present specification further provides an apparatus for identifying an illegal behavior in an unattended scenario, wherein the apparatus for identifying an illegal behavior in the unattended scenario includes: a processor And a memory for storing machine executable instructions.
  • the processor and the memory are usually connected to each other by an internal bus.
  • the device may also include an external interface to enable communication with other devices or components.
  • the processor by reading and executing the machine-executable instructions stored in the memory corresponding to the identification logic of the illegal behavior in the unattended scene, the processor is caused to:
  • the limb data includes: the mouth position data of the user;
  • the feature data includes: object position data of the object;
  • the processor is caused to: if the user is identified to have an illegal behavior based on the limb data and the feature data of the object:
  • the limb data further includes: the mouth behavior data of the user;
  • the processor is caused to: if the user is identified to have an illegal behavior based on the limb data and the feature data of the object:
  • the user is determined to have an illegal behavior when the distance of the mouth from the object is within a distance threshold for a predetermined length of time and the user's mouth behavior data matches a predetermined mouth motion pattern.
  • the feature data further includes: attribute category data of the object;
  • the processor is caused to: if the user is identified to have an illegal behavior based on the limb data and the feature data of the object:
  • the user is determined to have an illegal behavior when the distance of the mouth from the object is within a distance threshold for a predetermined length of time and the attribute category of the object belongs to an edible category.
  • the feature data further includes: image data of an outer surface of the object;
  • the processor is caused to: if the user is identified to have an illegal behavior based on the limb data and the feature data of the object:
  • the user is determined to have an illegal behavior when the distance of the mouth from the object is within a distance threshold for a predetermined length of time and the outer surface image of the object does not match the default outer surface image.
  • the limb data includes: behavior data of the limbs of the user;
  • the feature data includes: oscillating data of the object
  • the processor is caused to: if the user is identified to have an illegal behavior based on the limb data and the feature data of the object:
  • the oscillating data is collected by a sensor disposed on the object.
  • the processor is further caused to:
  • the processor is further caused to:
  • the illegal behavior is broadcasted in the unattended scene.
  • the processor is further caused to:
  • the amount of resources corresponding to the object is added to the bill of the user.
  • the present specification further provides a computer readable storage medium having a computer program stored thereon, the program being executed by the processor Implement the following steps:
  • the limb data includes: the mouth position data of the user;
  • the feature data includes: object position data of the object;
  • the identifying, according to the limb data and the feature data of the object, whether the user has an illegal behavior including:
  • the limb data further includes: the mouth behavior data of the user;
  • the identifying, according to the limb data and the feature data of the object, whether the user has an illegal behavior including:
  • the user is determined to have an illegal behavior when the distance of the mouth from the object is within a distance threshold for a predetermined length of time and the user's mouth behavior data matches a predetermined mouth motion pattern.
  • the feature data further includes: attribute category data of the object;
  • the identifying, according to the limb data and the feature data of the object, whether the user has an illegal behavior including:
  • the user is determined to have an illegal behavior when the distance of the mouth from the object is within a distance threshold for a predetermined length of time and the attribute category of the object belongs to an edible category.
  • the feature data further includes: image data of an outer surface of the object;
  • the identifying, according to the limb data and the feature data of the object, whether the user has an illegal behavior including:
  • the user is determined to have an illegal behavior when the distance of the mouth from the object is within a distance threshold for a predetermined length of time and the outer surface image of the object does not match the default outer surface image.
  • the limb data includes: behavior data of the limbs of the user;
  • the feature data includes: oscillating data of the object
  • the identifying, according to the limb data and the feature data of the object, whether the user has an illegal behavior including:
  • the oscillating data is collected by a sensor disposed on the object.
  • it also includes:
  • it also includes:
  • the illegal behavior is broadcasted in the unattended scene.
  • it also includes:
  • the amount of resources corresponding to the object is added to the bill of the user.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Theoretical Computer Science (AREA)
  • General Health & Medical Sciences (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Multimedia (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Human Computer Interaction (AREA)
  • Psychiatry (AREA)
  • Social Psychology (AREA)
  • Geometry (AREA)
  • Data Mining & Analysis (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Image Analysis (AREA)
  • Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)

Abstract

说明书披露一种无人值守场景中非法行为的识别方法和装置。该方法包括：采集位于无人值守场景中的用户的肢体数据；采集所述无人值守场景中物体的特征数据；根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为。

Description

无人值守场景中非法行为的识别方法和装置 技术领域
本说明书涉及互联网技术领域，尤其涉及一种无人值守场景中非法行为的识别方法和装置。
背景技术
随着科技的发展，无人值守场景越来越多的应用到日常生活中，例如，无人超市、无人健身房、无人KTV等。然而，由于无人看管，无人值守场景也面临着一些非法行为，例如：在无人超市中吃喝后不付款离去、破坏无人健身房中的健身设施等。因此，需要在无人值守场景中提供一种有效的非法行为识别方案。
发明内容
有鉴于此，本说明书提供一种无人值守场景中非法行为的识别方法和装置。
具体地，本说明书是通过如下技术方案实现的：
一种无人值守场景中非法行为的识别方法，包括：
采集位于无人值守场景中的用户的肢体数据；
采集所述无人值守场景中物体的特征数据；
根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为。
一种无人值守场景中非法行为的识别装置，包括：
第一采集单元，采集位于无人值守场景中的用户的肢体数据；
第二采集单元，采集所述无人值守场景中物体的特征数据；
行为识别单元，根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为。
一种无人值守场景中非法行为的识别装置，包括：
处理器；
用于存储机器可执行指令的存储器；
其中，通过读取并执行所述存储器存储的与无人值守场景中非法行为的识别逻辑对应的机器可执行指令，所述处理器被促使：
采集位于无人值守场景中的用户的肢体数据；
采集所述无人值守场景中物体的特征数据；
根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为。
由以上描述可以看出，本说明书可采集无人值守场景中用户的肢体数据和物体的特征数据，并综合用户和物体的实际情况对非法行为进行识别，从而实现对无人值守场景中非法行为的有效识别。
附图说明
图1是本说明书一示例性实施例示出的一种无人值守场景中非法行为的识别方法的流程示意图。
图2是本说明书一示例性实施例示出的另一种无人值守场景中非法行为的识别方法的流程示意图。
图3是本说明书一示例性实施例示出的另一种无人值守场景中非法行为的识别方法的流程示意图。
图4是本说明书一示例性实施例示出的一种用于无人值守场景中非法行为的识别装置的一结构示意图。
图5是本说明书一示例性实施例示出的一种无人值守场景中非法行为的识别装置的框图。
具体实施方式
这里将详细地对示例性实施例进行说明，其示例表示在附图中。下面的描述涉及附图时，除非另有表示，不同附图中的相同数字表示相同或相似的要素。以下示例性实施例中所描述的实施方式并不代表与本说明书相一致的所有实施方式。相反，它们仅是与如所附权利要求书中所详述的、本说明书的一些方面相一致的装置和方法的例子。
在本说明书使用的术语是仅仅出于描述特定实施例的目的，而非旨在限制本说明书。在本说明书和所附权利要求书中所使用的单数形式的“一种”、“所述”和“该”也旨 在包括多数形式，除非上下文清楚地表示其他含义。还应当理解，本文中使用的术语“和/或”是指并包含一个或多个相关联的列出项目的任何或所有可能组合。
应当理解，尽管在本说明书可能采用术语第一、第二、第三等来描述各种信息，但这些信息不应限于这些术语。这些术语仅用来将同一类型的信息彼此区分开。例如，在不脱离本说明书范围的情况下，第一信息也可以被称为第二信息，类似地，第二信息也可以被称为第一信息。取决于语境，如在此所使用的词语“如果”可以被解释成为“在……时”或“当……时”或“响应于确定”。
图1是本说明书一示例性实施例示出的一种无人值守场景中非法行为的识别方法的流程示意图。
所述无人值守场景中非法行为的识别方法可以应用在无人值守场景中的前端识别设备，例如：无人值守场景中部署的具有处理功能的识别设备；所述识别方法还可以应用在无人值守场景中的后端识别设备，例如，后台识别设备等，本说明书对此不作特殊限制。
请参考图1，所述无人值守场景中非法行为的识别方法可以包括以下步骤：
步骤102，采集位于无人值守场景中的用户的肢体数据。
在本实施例中，可在用户进入无人值守场景后开始进行所述肢体数据的采集。
所述肢体数据可包括：用户的肢体行为数据和肢体位置数据，用户的肢体可包括：四肢、嘴部等部位。
在本实施例中，可通过机器视觉技术采集无人值守场景中用户的所述肢体数据，当然，也可以通过其他技术采集所述肢体数据，本说明书对此不作特殊限制。
步骤104，采集所述无人值守场景中物体的特征数据。
在本实施例中，也可在用户进入无人值守场景后开始进行所述特征数据的采集。
其中，所述物体通常是无人值守场景中的物体，例如，无人超市中的货品、无人超市中摆放货品的货柜、无人健身房中的健身器材等。
所述特征数据可包括：所述物体的位置数据、所述物体的外表面图像、所述物体中设置的传感器采集到的震荡数据等。
在本实施例中，也可通过机器视觉技术采集无人值守场景中物体的所述特征数据，当然，也可以通过传感器等其他方式采集所述特征数据，本说明书对此不作特殊限制。
步骤106，根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为。
在本实施例中，可对采集到的所述肢体数据和所述特征数据进行分析，以判断无人值守场景中的用户是否存在非法行为，例如，破坏物品等。
由以上描述可以看出，本说明书可采集无人值守场景中用户的肢体数据和物体的特征数据，并综合用户和物体的实际情况对非法行为进行识别，从而实现对无人值守场景中非法行为的有效识别。
下面以无人值守场景是无人超市、以非法行为分别是非法吃喝、破坏货柜为例对本说明书的实现过程进行具体描述。
在无人超市的场景下，可以在用户进入无人超市后开始进行其肢体数据和物体特征数据的采集，以判断所述用户是否存在非法行为，并可在所述用户离开无人超市后停止对所述肢体数据和特征数据的采集。
其中，所述用户进入和离开无人超市的判别方法是多种多样的。
例如，可在用户从外部触发开启无人超市的大门后确定所述用户进入无人超市，可在所述用户从内部触发开启无人超市的大门后确定所述用户离开无人超市。
又例如，可在无人超市内的识别装置识别到活人后确定用户进入无人超市，可在无人超市内的识别装置无法识别到活人后确定所述用户离开无人超市。
上述触发无人超市的大门开启、识别装置对活人进行识别的方案都可以参照相关技术，本实施例在此不再一一赘述。
除上述方案之外，还可以采用其他方式确定用户是否进入或者离开无人超市，本实施例对此不作特殊限制。
一、非法行为是非法吃喝
在无人超市中，当用户进入无人超市后，可能会在超市内吃喝，在吃喝完毕后离去，给无人超市带来经济损失。
针对非法吃喝的非法行为，请参考图2，本实施例提供的识别方法可以包括以下步骤：
步骤202，采集位于无人超市中用户的嘴部位置数据。
在本实施例中，可以先采集用户的人脸特征，然后识别出用户的嘴部，进而得到用户的嘴部位置数据。例如，可在识别出用户的嘴部后，根据用户嘴部与采集装置之间的相对位置关系以及采集装置的位置数据计算出所述用户嘴部的位置数据。当然，在实际应用中，也可以采用其他方式采集得到用户的嘴部位置数据，本实施例对此不作特殊限制。
所述嘴部位置数据可以为嘴部的经纬度数据。
所述嘴部位置数据也可以是嘴部在该无人超市中的相对位置数据。例如，可以以无人超市内的某个空间点(例如，无人超市的墙脚)作为原点建立空间直角坐标系，然后将用户嘴部在该空间直角坐标系的坐标作为上述嘴部位置数据，本说明书对此不作特殊限制。
步骤204，采集所述无人超市中物体的物体位置数据。
在本实施例中，在对非法吃喝进行识别时，所述物体通常是无人超市中售卖的货品。
在本实施例中，若货品上设置有RFID(Radio Frequency Identification，射频识别)标签，则可以通过RFID定位技术确定货品位置数据。若货品上未设置RFID标签，可以通过其他定位技术确定货品位置数据，本实施例对此不作特殊限制。
在本实施例中，由于无人超市售卖的货品通常较多，所以可以预先保存各个货品的初始位置，例如，可将货品在货架上的位置作为初始位置。当某一货品的位置不在其初始位置时，可以执行本实施例的后续步骤。
在本实施例中，与用户的嘴部位置数据类似，所述物体位置数据可以为经纬度数据，也可以为该物体在无人超市中的相对位置数据，本实施例对此不作特殊限制。
步骤206，根据所述嘴部位置数据和所述物体位置数据计算用户的嘴部与所述物体的距离。
在本实施例中，在计算用户的嘴部与物体的距离时，通常采用同一时间点采集到的所述嘴部位置数据和所述物体位置数据，以确保上述距离的准确性。
步骤208，当嘴部与所述物体的距离在距离阈值内的时长达到预定时长时，确定所述用户存在非法吃喝行为。
在本实施例中，所述距离阈值可以预先设置，例如：20cm、30cm等。
当用户的嘴部与货品的距离在上述距离阈值内时，通常说明用户的嘴部与货品的距 离较近，用户可能是在闻货品的气味儿，用户也可能在偷吃或偷喝货品。
若用户想要闻货品的气味儿，那么用户的嘴部与货品的距离在上述距离阈值内的时长通常较短；若用户在偷吃或偷喝货品，那么用户的嘴部与货品的距离在上述距离阈值内的时长通常较长，因此，可以在用户嘴部与货品的距离在距离阈值内的时长达到预定时长时，确定用户在偷吃或偷喝，即用户存在非法吃喝行为。
其中，上述预定时长也可以预先设置，例如：5秒、10秒等。
由以上描述可以看出，本实施例可在无人值守场景中用户的嘴部与货品的距离在距离阈值内的时长达到预定时长时，确定用户存在非法吃喝行为，进而实现对无人值守场景中非法吃喝行为的识别。
可选的，为避免误判、提高非法吃喝行为识别的准确率，还可以采集所述用户的嘴部行为数据作为其肢体数据，并判断该嘴部行为数据是否匹配预定的嘴部运动规律，例如：咀嚼规律、吞咽规律等。
若采集到的嘴部行为数据匹配预定的嘴部运动规律、且用户嘴部与货品的距离在距离阈值内的时长达到预定时长，则可以确定用户在偷吃或偷喝，即用户存在非法吃喝行为。
若采集到的嘴部行为数据不匹配预定的嘴部运动规律，则可以确定用户不存在非法吃喝行为。
其中，用户的嘴部行为数据与预定的嘴部运动规律匹配与否的判断可以在图2所示的步骤206之前，也可以在步骤206之后，还可以与步骤206同时进行，本实施例对此不作特殊限制。
可选的，为避免误判、提高非法吃喝行为识别的准确率，还可以采集货品的属性类别数据作为所述货品的特征数据。该属性类别数据通常是指货品的品类划分，例如：可食用、不可食用等。
若用户在无人超市中非法吃喝，那么用户选择的货品必定为可食用类别的货品。因此，可在用户的嘴部与货品的距离在距离阈值内的时长达到预定时长、且货品的属性类别属于可食用类别时，确定用户在偷吃或偷喝。例如，货品的属性类别是瓶装饮用水，且用户与该瓶装饮用水的距离在距离阈值内的时长达到预定时长，则可以确定用户在偷喝该瓶装饮用水。
若用户的嘴部与货品的距离在距离阈值内的时长达到预定时长、但货品的属性类别不属于可食用类别时，可以确定用户没有在偷吃或偷喝。例如，货品的属性类别是家居用品，虽然用户与该家居用品的距离在距离阈值内的时长达到预定时长，则也可以确定用户不存在非法吃喝行为。
其中，对货品的属性类别的判断可以在图2所示的步骤206之前，也可以在步骤206之后，还可以与步骤206同时进行，本实施例对此不作特殊限制。
可选的，为避免误判、提高非法吃喝行为识别的准确率，还可以采集货品的外表面图像数据作为货品的特征数据，该外表面图像数据通常可以理解为货品的外包装图像。
考虑到无人超市中的某些可食用货品需要拆开包装后食用，还可以判断货品的外表面图像数据是否匹配其缺省的外表面图像。其中，该缺省的外表面图像可被预先存储。若货品的外表面图像数据匹配其缺省的外表面图像，则可以说明货品的外包装没有被破坏；若货品的外表面图像数据不匹配其缺省的外表面图像，则可以说明货品的外包装已经被破坏。
在本例中，可以在用户的嘴部与货品的距离在距离阈值内的时长达到预定时长、且货品的外包装被破坏时，确定所述用户存在非法行为。
其中，对货品的外表面图像数据的匹配可以在图2所示的步骤206之前，也可以在步骤206之后，还可以与步骤206同时进行，本实施例对此不作特殊限制。
在实际应用中，在对用户是否存在非法吃喝行为进行识别时，可以结合上述各种识别方案，也可以采用其他的识别方案，本说明书对此不作特殊限制。
二、非法行为是破坏货柜
在无人超市中，当用户进入无人超市后，可能会破坏无人超市的货品等物品，例如，用脚踢货柜等，给无人超市带来经济损失。
针对非法破坏货柜等非法行为，请参考图3，本实施例提供的识别方法可以包括以下步骤：
步骤302，采集位于无人超市中用户的四肢行为数据。
在本实施例中，所述四肢行为数据可以包括：手臂行为数据、腿部行为数据等。
在本实施例中，可以采集用户在一段时间内的四肢行为数据，以供分析用户的四肢运动情况。
步骤304，采集无人超市货柜的震荡数据。
在本实施例中，可在超市的货柜上设置传感器，例如：陀螺仪传感器、加速度传感器等，以采集货柜的震荡数据。
当然，若货柜上未设置传感器，也可以通过机器视觉等其他技术采集货柜的震荡数据。
基于采集到的震荡数据，可以确定货柜是否发生震荡。
步骤306，当所述用户的四肢行为数据匹配预定的四肢运动规律、且所述货柜发生震荡时，确定所述用户存在非法行为。
基于前述步骤302，在采集到用户的四肢行为数据后，可以判断所述四肢行为数据是否匹配预定的四肢运动规律，例如：“砸”规律、“踢”规律、“踹”规律等。
若所述四肢行为数据匹配所述预定的四肢运动规律，说明用户可能在破坏无人超市内的物品，可继续进行判断。
若所述四肢行为数据不匹配所述预定的四肢运动规律，则可以说明用户不存在破坏无人超市内物品的非法行为。
基于前述步骤304，在采集到货柜的震荡数据后，可根据该震荡数据判断货柜是否发生震荡。
若货柜发生震荡、且用户的四肢行为数据匹配上述预定的四肢运动规律，则可以确定用户存在破坏货柜的非法行为。
若货柜未发生震荡，那么可以确定用户不存在破坏货柜的非法行为。
由以上描述可以看出，本实施例可在无人值守场景中用户的四肢行为数据匹配预定的破坏类四肢运动规律、且物体发生震荡时，确定用户存在破坏物体的非法行为，进而实现对无人值守场景中非法破坏行为的识别。
可选的，在另一个例子中，在进行非法破坏行为的识别时，还可以计算用户匹配上述四肢运动规律的肢体与上述物体之间的距离，若该距离在预定的范围内，即距离较近时，可以确定用户存在非法破坏物体的行为。
可选的，在另一个例子中，在无人超市中，除货柜外，用户还可能破坏无人超市售卖的货品，例如：撕毁货品的外包装等。因此，在进行非法破坏行为的识别时，也可以采集物体的外表面图像数据，然后判断采集到的外表面图像数据和物体的缺省外表 面图像是否匹配，以判断物体的外包装是否有被破坏。
针对无人超市的货柜、无人健身房的健身器材等物体，也可以采集其外表面图像数据与其缺省外表面图像进行匹配，本实施例对此不作特殊限制。然而，由于货柜、健身器材等物体通常较为结实，即便用户对其进行破坏，也不一定会导致其外表面图像有变化，为提高识别准确性，通常还需要采集其他数据进行识别。
在实际应用中，在对用户是否存在非法破坏行为进行识别时，可以结合上述各种识别方案，也可以采用其他的识别方案，本说明书对此不作特殊限制。
此外，在实现上述各种识别方案时，通常需要开发人员先建立对应的行为识别模型，然后通过对该模型的不断训练以完善该行为识别模型。在该行为识别模型完善之前，若根据用户的肢体数据和物体的特征数据无法判别用户到底是否存在非法行为，也可以输出判断请求给相关的工作人员，由工作人员人工进行判断，并将判断结果返回，以完善上述行为识别模型。当然，在根据用户的肢体数据和物体的特征数据无法判别用户到底是否存在非法行为时，也可以继续进行上述肢体数据和特征数据的采集，并继续进行识别，若到达一定的时长仍无法识别，则也可以输出判断请求给相关的工作人员，以请求人工介入。
在识别出用户存在非法行为后，可以通过多种不同的方式对用户的非法行为进行告警、惩戒。
在一个例子中，当用户存在非法行为时，可以对用户进行信用惩戒。
具体而言，用户在进入无人值守场景之前，通常要经过身份认证，例如，在进入无人超市前，通过扫描二维码以供无人超市获取到用户的账号信息。
若用户存在非法行为，则可以负向更新用户的用户账号的信用信息，例如，降低其信用评分等。
在另一个例子中，当用户存在非法行为时，可以在所述无人值守场景中播报所述非法行为。例如，可以通过音频播报该用户的非法行为，也可以通过视频回放该用户的非法行为，以进行警示。
在另一个例子中，当用户存在非法行为时，还可以将该非法行为针对的物体对应的资源量添加到所述用户的账单中。其中，所述资源量通常是指该物体的价值。
举例来说，假设用户在无人超市中非法食用薯片，那么可将薯片的金额添加到 用户的账单中，即便用户在吃完薯片后将包装丢弃在无人超市内，也会向用户收取对应的费用。
又假设，用户在无人超市内破坏货柜，那么可以将破坏货柜行为对应的罚款金额添加到用户的账单中，以实现对用户进行惩戒的目的。
当然，在实际应用中，当用户存在非法行为时，还可以采用其他方式对用户进行惩戒，例如，将用户的非法行为在行业内进行通告等，本说明书对此不作特殊限制。
与前述无人值守场景中非法行为的识别方法的实施例相对应，本说明书还提供了无人值守场景中非法行为的识别装置的实施例。
本说明书无人值守场景中非法行为的识别装置的实施例可以应用在识别设备上。装置实施例可以通过软件实现，也可以通过硬件或者软硬件结合的方式实现。以软件实现为例，作为一个逻辑意义上的装置，是通过其所在识别设备的处理器将非易失性存储器中对应的计算机程序指令读取到内存中运行形成的。从硬件层面而言，如图4所示，为本说明书无人值守场景中非法行为的识别装置所在识别设备的一种硬件结构图，除了图4所示的处理器、内存、网络接口、以及非易失性存储器之外，实施例中装置所在的识别设备通常根据该识别设备的实际功能，还可以包括其他硬件，对此不再赘述。
图5是本说明书一示例性实施例示出的一种无人值守场景中非法行为的识别装置的框图。
请参考图5，所述无人值守场景中非法行为的识别装置400可以应用在前述图4所示的识别设备中，包括有：第一采集单元401、第二采集单元402、行为识别单元403、信用惩戒单元404、播报惩戒单元405以及资源惩戒单元406。
其中，第一采集单元401，采集位于无人值守场景中的用户的肢体数据；
第二采集单元402，采集所述无人值守场景中物体的特征数据；
行为识别单元403，根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为。
可选的，所述肢体数据包括：所述用户的嘴部位置数据；
所述特征数据包括：所述物体的物体位置数据；
所述行为识别单元403，根据所述嘴部位置数据和所述物体位置数据计算所述用户的嘴部与所述物体的距离；当嘴部与所述物体的距离在距离阈值内的时长达到预定时 长时，确定所述用户存在非法行为。
可选的，所述肢体数据还包括：所述用户的嘴部行为数据；
所述行为识别单元403，在嘴部与所述物体的距离在距离阈值内的时长达到预定时长、且所述用户的嘴部行为数据匹配预定的嘴部运动规律时，确定所述用户存在非法行为。
可选的，所述特征数据还包括：所述物体的属性类别数据；
所述行为识别单元403，在嘴部与所述物体的距离在距离阈值内的时长达到预定时长、且所述物体的属性类别属于可食用类别时，确定所述用户存在非法行为。
可选的，所述特征数据还包括：所述物体的外表面图像数据；
所述行为识别单元403，在嘴部与所述物体的距离在距离阈值内的时长达到预定时长、且所述物体的外表面图像与缺省外表面图像不匹配时，确定所述用户存在非法行为。
可选的，所述肢体数据包括：所述用户的四肢行为数据；
所述特征数据包括：所述物体的震荡数据；
所述行为识别单元403，在所述用户的四肢行为数据匹配预定的四肢运动规律、且所述物体发生震荡时，确定所述用户存在非法行为。
可选的，所述震荡数据由所述物体上设置的传感器采集。
信用惩戒单元404，当所述用户存在非法行为时，对所述用户进行信用惩戒。
播报惩戒单元405，当所述用户存在非法行为时，在所述无人值守场景中播报所述非法行为。
资源惩戒单元406，当所述用户存在非法行为时，将所述物体对应的资源量添加到所述用户的账单中。
上述装置中各个单元的功能和作用的实现过程具体详见上述方法中对应步骤的实现过程，在此不再赘述。
对于装置实施例而言，由于其基本对应于方法实施例，所以相关之处参见方法实施例的部分说明即可。以上所描述的装置实施例仅仅是示意性的，其中所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者 也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部模块来实现本说明书方案的目的。本领域普通技术人员在不付出创造性劳动的情况下，即可以理解并实施。
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，计算机的具体形式可以是个人计算机、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、电子邮件收发设备、游戏控制台、平板计算机、可穿戴设备或者这些设备中的任意几种设备的组合。
与前述无人值守场景中非法行为的识别方法的实施例相对应，本说明书还提供一种无人值守场景中非法行为的识别装置，该无人值守场景中非法行为的识别装置包括：处理器以及用于存储机器可执行指令的存储器。其中，处理器和存储器通常借由内部总线相互连接。在其他可能的实现方式中，所述设备还可能包括外部接口，以能够与其他设备或者部件进行通信。
在本实施例中，通过读取并执行所述存储器存储的与无人值守场景中非法行为的识别逻辑对应的机器可执行指令，所述处理器被促使：
采集位于无人值守场景中的用户的肢体数据；
采集所述无人值守场景中物体的特征数据；
根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为。
可选的，所述肢体数据包括：所述用户的嘴部位置数据；
所述特征数据包括：所述物体的物体位置数据；
在根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为时，所述处理器被促使：
根据所述嘴部位置数据和所述物体位置数据计算所述用户的嘴部与所述物体的距离；
当嘴部与所述物体的距离在距离阈值内的时长达到预定时长时，确定所述用户存在非法行为。
可选的，所述肢体数据还包括：所述用户的嘴部行为数据；
在根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为 时，所述处理器被促使：
当嘴部与所述物体的距离在距离阈值内的时长达到预定时长、且所述用户的嘴部行为数据匹配预定的嘴部运动规律时，确定所述用户存在非法行为。
可选的，所述特征数据还包括：所述物体的属性类别数据；
在根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为时，所述处理器被促使：
当嘴部与所述物体的距离在距离阈值内的时长达到预定时长、且所述物体的属性类别属于可食用类别时，确定所述用户存在非法行为。
可选的，所述特征数据还包括：所述物体的外表面图像数据；
在根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为时，所述处理器被促使：
当嘴部与所述物体的距离在距离阈值内的时长达到预定时长、且所述物体的外表面图像与缺省外表面图像不匹配时，确定所述用户存在非法行为。
可选的，所述肢体数据包括：所述用户的四肢行为数据；
所述特征数据包括：所述物体的震荡数据；
在根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为时，所述处理器被促使：
当所述用户的四肢行为数据匹配预定的四肢运动规律、且所述物体发生震荡时，确定所述用户存在非法行为。
可选的，所述震荡数据由所述物体上设置的传感器采集。
可选的，通过读取并执行所述存储器存储的与无人值守场景中非法行为的识别逻辑对应的机器可执行指令，所述处理器还被促使：
当所述用户存在非法行为时，对所述用户进行信用惩戒。
可选的，通过读取并执行所述存储器存储的与无人值守场景中非法行为的识别逻辑对应的机器可执行指令，所述处理器还被促使：
当所述用户存在非法行为时，在所述无人值守场景中播报所述非法行为。
可选的，通过读取并执行所述存储器存储的与无人值守场景中非法行为的识别 逻辑对应的机器可执行指令，所述处理器还被促使：
当所述用户存在非法行为时，将所述物体对应的资源量添加到所述用户的账单中。
与前述无人值守场景中非法行为的识别方法的实施例相对应，本说明书还提供一种计算机可读存储介质，所述计算机可读存储介质上存储有计算机程序，该程序被处理器执行时实现以下步骤：
采集位于无人值守场景中的用户的肢体数据；
采集所述无人值守场景中物体的特征数据；
根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为。
可选的，所述肢体数据包括：所述用户的嘴部位置数据；
所述特征数据包括：所述物体的物体位置数据；
所述根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为，包括：
根据所述嘴部位置数据和所述物体位置数据计算所述用户的嘴部与所述物体的距离；
当嘴部与所述物体的距离在距离阈值内的时长达到预定时长时，确定所述用户存在非法行为。
可选的，所述肢体数据还包括：所述用户的嘴部行为数据；
所述根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为，包括：
当嘴部与所述物体的距离在距离阈值内的时长达到预定时长、且用户的嘴部行为数据匹配预定的嘴部运动规律时，确定所述用户存在非法行为。
可选的，所述特征数据还包括：所述物体的属性类别数据；
所述根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为，包括：
当嘴部与所述物体的距离在距离阈值内的时长达到预定时长、且所述物体的属性类别属于可食用类别时，确定所述用户存在非法行为。
可选的，所述特征数据还包括：所述物体的外表面图像数据；
所述根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为，包括：
当嘴部与所述物体的距离在距离阈值内的时长达到预定时长、且所述物体的外表面图像与缺省外表面图像不匹配时，确定所述用户存在非法行为。
可选的，所述肢体数据包括：所述用户的四肢行为数据；
所述特征数据包括：所述物体的震荡数据；
所述根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为，包括：
当所述用户的四肢行为数据匹配预定的四肢运动规律、且所述物体发生震荡时，确定所述用户存在非法行为。
可选的，所述震荡数据由所述物体上设置的传感器采集。
可选的，还包括：
当所述用户存在非法行为时，对所述用户进行信用惩戒。
可选的，还包括：
当所述用户存在非法行为时，在所述无人值守场景中播报所述非法行为。
可选的，还包括：
当所述用户存在非法行为时，将所述物体对应的资源量添加到所述用户的账单中。
上述对本说明书特定实施例进行了描述。其它实施例在所附权利要求书的范围内。在一些情况下，在权利要求书中记载的动作或步骤可以按照不同于实施例中的顺序来执行并且仍然可以实现期望的结果。另外，在附图中描绘的过程不一定要求示出的特定顺序或者连续顺序才能实现期望的结果。在某些实施方式中，多任务处理和并行处理也是可以的或者可能是有利的。
以上所述仅为本说明书的较佳实施例而已，并不用以限制本说明书，凡在本说明书的精神和原则之内，所做的任何修改、等同替换、改进等，均应包含在本说明书保护的范围之内。

Claims (21)

  1. 一种无人值守场景中非法行为的识别方法，包括：
    采集位于无人值守场景中的用户的肢体数据；
    采集所述无人值守场景中物体的特征数据；
    根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为。
  2. 根据权利要求1所述的方法，
    所述肢体数据包括：所述用户的嘴部位置数据；
    所述特征数据包括：所述物体的物体位置数据；
    所述根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为，包括：
    根据所述嘴部位置数据和所述物体位置数据计算所述用户的嘴部与所述物体的距离；
    当嘴部与所述物体的距离在距离阈值内的时长达到预定时长时，确定所述用户存在非法行为。
  3. 根据权利要求2所述的方法，
    所述肢体数据还包括：所述用户的嘴部行为数据；
    所述根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为，包括：
    当嘴部与所述物体的距离在距离阈值内的时长达到预定时长、且所述用户的嘴部行为数据匹配预定的嘴部运动规律时，确定所述用户存在非法行为。
  4. 根据权利要求2所述的方法，
    所述特征数据还包括：所述物体的属性类别数据；
    所述根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为，包括：
    当嘴部与所述物体的距离在距离阈值内的时长达到预定时长、且所述物体的属性类别属于可食用类别时，确定所述用户存在非法行为。
  5. 根据权利要求2所述的方法，
    所述特征数据还包括：到的物体的外表面图像数据；
    所述根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为，包括：
    当嘴部与所述物体的距离在距离阈值内的时长达到预定时长、且所述物体的外表面 图像与缺省外表面图像不匹配时，确定所述用户存在非法行为。
  6. 根据权利要求1所述的方法，
    所述肢体数据包括：所述用户的四肢行为数据；
    所述特征数据包括：所述物体的震荡数据；
    所述根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为，包括：
    当所述用户的四肢行为数据匹配预定的四肢运动规律、且所述物体发生震荡时，确定所述用户存在非法行为。
  7. 根据权利要求6所述的方法，
    所述震荡数据由所述物体上设置的传感器采集。
  8. 根据权利要求1所述的方法，还包括：
    当所述用户存在非法行为时，对所述用户进行信用惩戒。
  9. 根据权利要求1所述的方法，还包括：
    当所述用户存在非法行为时，在所述无人值守场景中播报所述非法行为。
  10. 根据权利要求1所述的方法，还包括：
    当所述用户存在非法行为时，将所述物体对应的资源量添加到所述用户的账单中。
  11. 一种无人值守场景中非法行为的识别装置，包括：
    第一采集单元，采集位于无人值守场景中的用户的肢体数据；
    第二采集单元，采集所述无人值守场景中物体的特征数据；
    行为识别单元，根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为。
  12. 根据权利要求11所述的装置，
    所述肢体数据包括：所述用户的嘴部位置数据；
    所述特征数据包括：所述物体的物体位置数据；
    所述行为识别单元，根据所述嘴部位置数据和所述物体位置数据计算所述用户的嘴部与所述物体的距离；当嘴部与所述物体的距离在距离阈值内的时长达到预定时长时，确定所述用户存在非法行为。
  13. 根据权利要求12所述的装置，
    所述肢体数据还包括：所述用户的嘴部行为数据；
    所述行为识别单元，在嘴部与所述物体的距离在距离阈值内的时长达到预定时长、且所述用户的嘴部行为数据匹配预定的嘴部运动规律时，确定所述用户存在非法行为。
  14. 根据权利要求12所述的装置，
    所述特征数据还包括：所述物体的属性类别数据；
    所述行为识别单元，在嘴部与所述物体的距离在距离阈值内的时长达到预定时长、且所述物体的属性类别属于可食用类别时，确定所述用户存在非法行为。
  15. 根据权利要求12所述的装置，
    所述特征数据还包括：所述物体的外表面图像数据；
    所述行为识别单元，在嘴部与所述物体的距离在距离阈值内的时长达到预定时长、且所述物体的外表面图像与缺省外表面图像不匹配时，确定所述用户存在非法行为。
  16. 根据权利要求11所述的装置，
    所述肢体数据包括：所述用户的四肢行为数据；
    所述特征数据包括：所述物体的震荡数据；
    所述行为识别单元，在所述用户的四肢行为数据匹配预定的四肢运动规律、且所述物体发生震荡时，确定所述用户存在非法行为。
  17. 根据权利要求16所述的装置，
    所述震荡数据由所述物体上设置的传感器采集。
  18. 根据权利要求11所述的装置，还包括：
    信用惩戒单元，当所述用户存在非法行为时，对所述用户进行信用惩戒。
  19. 根据权利要求11所述的装置，还包括：
    播报惩戒单元，当所述用户存在非法行为时，在所述无人值守场景中播报所述非法行为。
  20. 根据权利要求11所述的装置，还包括：
    资源惩戒单元，当所述用户存在非法行为时，将所述物体对应的资源量添加到所述用户的账单中。
  21. 一种无人值守场景中非法行为的识别装置，包括：
    处理器；
    用于存储机器可执行指令的存储器；
    其中，通过读取并执行所述存储器存储的与无人值守场景中非法行为的识别逻辑对应的机器可执行指令，所述处理器被促使：
    采集位于无人值守场景中的用户的肢体数据；
    采集所述无人值守场景中物体的特征数据；
    根据所述肢体数据和所述物体的特征数据识别所述用户是否存在非法行为。
PCT/CN2018/100977 2017-11-03 2018-08-17 无人值守场景中非法行为的识别方法和装置 Ceased WO2019085589A1 (zh)

Priority Applications (5)

Application Number Priority Date Filing Date Title
KR1020207010861A KR102256895B1 (ko) 2017-11-03 2018-08-17 무인 시나리오에서의 불법 행위를 인식하기 위한 방법 및 장치
KR1020217014172A KR102429739B1 (ko) 2017-11-03 2018-08-17 무인 시나리오에서의 불법 행위를 인식하기 위한 방법 및 장치
SG11202003011TA SG11202003011TA (en) 2017-11-03 2018-08-17 Method and apparatus for recognizing illegal behavior in unattended scenario
US16/796,616 US10783362B2 (en) 2017-11-03 2020-02-20 Method and apparatus for recognizing illegal behavior in unattended scenario
US16/943,639 US10990813B2 (en) 2017-11-03 2020-07-30 Method and apparatus for recognizing illegal behavior in unattended scenario

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201711069063.2 2017-11-03
CN201711069063.2A CN108021865B (zh) 2017-11-03 2017-11-03 无人值守场景中非法行为的识别方法和装置

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US16/796,616 Continuation US10783362B2 (en) 2017-11-03 2020-02-20 Method and apparatus for recognizing illegal behavior in unattended scenario

Publications (1)

Publication Number Publication Date
WO2019085589A1 true WO2019085589A1 (zh) 2019-05-09

Family

ID=62080436

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/100977 Ceased WO2019085589A1 (zh) 2017-11-03 2018-08-17 无人值守场景中非法行为的识别方法和装置

Country Status (6)

Country Link
US (2) US10783362B2 (zh)
KR (2) KR102256895B1 (zh)
CN (2) CN111507216A (zh)
SG (1) SG11202003011TA (zh)
TW (1) TWI686776B (zh)
WO (1) WO2019085589A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110427858A (zh) * 2019-07-26 2019-11-08 广州利科科技有限公司 一种结合图像的人物行为识别轨迹方法
CN113011449A (zh) * 2019-12-20 2021-06-22 中移(上海)信息通信科技有限公司 行为确定方法、装置、设备及存储介质

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10113877B1 (en) * 2015-09-11 2018-10-30 Philip Raymond Schaefer System and method for providing directional information
CN111507216A (zh) 2017-11-03 2020-08-07 阿里巴巴集团控股有限公司 无人值守场景中非法行为的识别方法和装置
CN109146499A (zh) * 2018-09-05 2019-01-04 深圳正品创想科技有限公司 基于区块链的用户信用调整方法及其装置、区块链节点
WO2020065839A1 (ja) * 2018-09-27 2020-04-02 株式会社オプティム 物体状況判断システム、物体状況判断方法、およびプログラム
CN110826506A (zh) * 2019-11-11 2020-02-21 上海秒针网络科技有限公司 目标行为的识别方法及装置
KR102465608B1 (ko) * 2020-10-26 2022-11-11 한국생산기술연구원 동작인식 기반 무인택배 관리시스템 및 이의 제어방법
US11771991B2 (en) * 2021-02-15 2023-10-03 Nintendo Co., Ltd. Non-transitory computer-readable storage medium having stored therein information processing program, information processing apparatus, information processing system, and information processing method
CN116030584B (zh) * 2023-01-17 2025-09-12 中为新智(成都)科技有限责任公司 一种实时视频流停车场车辆偷油事件检测方法及系统
CN117319010A (zh) * 2023-09-15 2023-12-29 中国银行股份有限公司 一种应用于边缘计算的访问方法、装置、设备及介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102881100A (zh) * 2012-08-24 2013-01-16 济南纳维信息技术有限公司 基于视频分析的实体店面防盗监控方法
CN106204240A (zh) * 2016-07-23 2016-12-07 山东华旗新能源科技有限公司 智慧购物管理系统
CN106981150A (zh) * 2017-04-19 2017-07-25 西安超嗨网络科技有限公司 一种超市智能防盗系统及方法
CN108021865A (zh) * 2017-11-03 2018-05-11 阿里巴巴集团控股有限公司 无人值守场景中非法行为的识别方法和装置

Family Cites Families (33)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2000293685A (ja) 1999-04-06 2000-10-20 Toyota Motor Corp シーン認識装置
US9159100B2 (en) * 2003-04-23 2015-10-13 Pitney Bowes Inc. System and method for dynamically managing business machine assets
GB2403363A (en) 2003-06-25 2004-12-29 Hewlett Packard Development Co Tags for automated image processing
TW200525449A (en) * 2004-01-29 2005-08-01 Lai Jin Ding Human body image recognition system
TW200533147A (en) * 2004-03-31 2005-10-01 Chong-Yu Yao Electronic recognition and confirmation system with burglarproof and safety secured function
US7639840B2 (en) 2004-07-28 2009-12-29 Sarnoff Corporation Method and apparatus for improved video surveillance through classification of detected objects
NL1026951C2 (nl) 2004-09-02 2006-03-09 Nedap Nv Elektronisch diefstaldetectiesysteem, alsmede een gegevensverwerkingssysteem en een werkwijze voor het tegengaan van diefstal van artikelen.
US7671741B2 (en) 2005-07-27 2010-03-02 Lax Michael R Anti-theft security device and perimeter detection system
US20080172335A1 (en) * 2007-01-11 2008-07-17 Chi-Chen Cheng User credit rating system to protect digital data
TW200837644A (en) * 2007-03-06 2008-09-16 Rf Matrix Inc A secure commerce and asset/supply chain management system
US7782207B2 (en) 2007-06-12 2010-08-24 Checkpoint Systems, Inc. Comprehensive theft security system
US8189905B2 (en) 2007-07-11 2012-05-29 Behavioral Recognition Systems, Inc. Cognitive model for a machine-learning engine in a video analysis system
US7920063B2 (en) 2007-08-13 2011-04-05 Wal-Mart Stores, Inc. RFID theft prevention system
JP4972491B2 (ja) 2007-08-20 2012-07-11 株式会社構造計画研究所 顧客動作判定システム
MY159289A (en) 2008-09-24 2016-12-30 Mimos Berhad A system and a method for identifying human behavioural intention based on an effective motion analysis
CN101571982B (zh) 2009-05-11 2011-04-13 宁波海视智能系统有限公司 一种判断视频监控范围内物品失窃的方法
CN202217347U (zh) * 2011-01-26 2012-05-09 北京海鑫智圣技术有限公司 银行atm机全方位智能监控设备
US9349234B2 (en) * 2012-03-14 2016-05-24 Autoconnect Holdings Llc Vehicle to vehicle social and business communications
US20150379238A1 (en) * 2012-06-14 2015-12-31 Medibotics Llc Wearable Imaging Device for Monitoring Food Consumption Using Gesture Recognition
CN102930250B (zh) * 2012-10-23 2015-11-18 西安理工大学 一种多尺度条件随机场模型的动作识别方法
KR101400730B1 (ko) * 2012-11-09 2014-05-29 재단법인대구경북과학기술원 지능형 냉장고 및 지능형 냉장고를 제어하는 방법
US20140180757A1 (en) 2012-12-20 2014-06-26 Wal-Mart Stores, Inc. Techniques For Recording A Consumer Shelf Experience
JP5356615B1 (ja) * 2013-02-01 2013-12-04 パナソニック株式会社 顧客行動分析装置、顧客行動分析システムおよび顧客行動分析方法
US10176456B2 (en) 2013-06-26 2019-01-08 Amazon Technologies, Inc. Transitioning items from a materials handling facility
US10268983B2 (en) 2013-06-26 2019-04-23 Amazon Technologies, Inc. Detecting item interaction and movement
CN103873825A (zh) * 2014-02-28 2014-06-18 北京航科威视光电信息技术有限公司 一种atm智能监控系统及方法
US10380877B2 (en) * 2014-08-05 2019-08-13 Overview Technologies, Inc. Community security system using intelligent information sharing
US9544551B2 (en) 2014-09-29 2017-01-10 Tyco Fire & Security Gmbh Store intelligence platform using proximity sensing
CN104616186B (zh) * 2014-11-05 2019-02-01 青岛海信智能商用系统股份有限公司 自助收款防损系统、自助结账防损系统和自助结账方法
DE102016113333B4 (de) * 2016-07-20 2020-06-25 Lsp Innovative Automotive Systems Gmbh System und Verfahren zum Sichern eines Fahrzeugs
CN106778179B (zh) * 2017-01-05 2021-07-09 南京大学 一种基于超声波唇语识别的身份认证方法
CN107146386B (zh) * 2017-05-05 2019-12-31 广东小天才科技有限公司 一种异常行为检测方法及装置、用户设备
JP6925879B2 (ja) * 2017-06-13 2021-08-25 任天堂株式会社 ゲームプログラム、情報処理装置、情報処理システム、および、ゲーム処理方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102881100A (zh) * 2012-08-24 2013-01-16 济南纳维信息技术有限公司 基于视频分析的实体店面防盗监控方法
CN106204240A (zh) * 2016-07-23 2016-12-07 山东华旗新能源科技有限公司 智慧购物管理系统
CN106981150A (zh) * 2017-04-19 2017-07-25 西安超嗨网络科技有限公司 一种超市智能防盗系统及方法
CN108021865A (zh) * 2017-11-03 2018-05-11 阿里巴巴集团控股有限公司 无人值守场景中非法行为的识别方法和装置

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110427858A (zh) * 2019-07-26 2019-11-08 广州利科科技有限公司 一种结合图像的人物行为识别轨迹方法
CN110427858B (zh) * 2019-07-26 2022-07-08 广州利科科技有限公司 一种结合图像的人物行为识别轨迹方法
CN113011449A (zh) * 2019-12-20 2021-06-22 中移(上海)信息通信科技有限公司 行为确定方法、装置、设备及存储介质

Also Published As

Publication number Publication date
US10990813B2 (en) 2021-04-27
CN111507216A (zh) 2020-08-07
US20200193149A1 (en) 2020-06-18
KR20210057838A (ko) 2021-05-21
TWI686776B (zh) 2020-03-01
KR102256895B1 (ko) 2021-05-31
KR20200054277A (ko) 2020-05-19
US20200356764A1 (en) 2020-11-12
US10783362B2 (en) 2020-09-22
CN108021865A (zh) 2018-05-11
TW201923705A (zh) 2019-06-16
KR102429739B1 (ko) 2022-08-04
CN108021865B (zh) 2020-02-21
SG11202003011TA (en) 2020-04-29

Similar Documents

Publication Publication Date Title
CN108021865B (zh) 无人值守场景中非法行为的识别方法和装置
US12394189B2 (en) Data recognition model construction apparatus and method for constructing data recognition model thereof, and data recognition apparatus and method for recognizing data thereof
US7925081B2 (en) Systems and methods for human body pose estimation
CN104487915B (zh) 维持扩增的连续性
JP2024528528A (ja) デジタル画像処理を用いたアイテム識別
US9832353B2 (en) Methods for encoding, decoding and interpreting auxiliary data in media signals
US10592687B2 (en) Method and system of enforcing privacy policies for mobile sensory devices
CN104424585A (zh) 播放方法及电子装置
CN112150349A (zh) 一种图像处理方法、装置、计算机设备及存储介质
WO2023070510A1 (en) Systems and methods for performing behavior detection and behavioral intervention
CN114360182A (zh) 一种智能告警方法、装置、设备及存储介质
CN110895691B (zh) 图像处理方法、装置及电子设备
HK1254017B (zh) 无人值守场景中非法行为的识别方法和装置
HK40034129A (zh) 无人值守场景中非法行为的识别方法和装置
HK1254017A1 (zh) 无人值守场景中非法行为的识别方法和装置
EP4373597A1 (en) Systems and methods for customized non-fungible tokens corresponding to an amusement park attraction
CN110187921A (zh) 唤醒设备的方法及装置
JP7750598B2 (ja) 振舞認識装置、振舞認識方法、及びプログラム
CN118747867B (zh) 一种基于生物特征识别的智能监控方法及系统
CN121963299A (zh) 基于姿态估计的行人跌倒检测方法及系统
WO2023217193A1 (zh) 机器人和机器人识别跌倒的方法
HK40013173A (zh) 唤醒设备的方法及装置
KR20160027503A (ko) 비콘 신호를 이용한 이미테이션 방지 방법 및 장치

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: 18873258

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 20207010861

Country of ref document: KR

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 18873258

Country of ref document: EP

Kind code of ref document: A1