WO2019165895A1 - 自动售货方法和系统以及自动售货装置和自动贩售机 - Google Patents

自动售货方法和系统以及自动售货装置和自动贩售机 Download PDF

Info

Publication number
WO2019165895A1
WO2019165895A1 PCT/CN2019/075147 CN2019075147W WO2019165895A1 WO 2019165895 A1 WO2019165895 A1 WO 2019165895A1 CN 2019075147 W CN2019075147 W CN 2019075147W WO 2019165895 A1 WO2019165895 A1 WO 2019165895A1
Authority
WO
WIPO (PCT)
Prior art keywords
user
shelf
information
item
vending
Prior art date
Application number
PCT/CN2019/075147
Other languages
English (en)
French (fr)
Inventor
张钟毓
陈宇
刘巍
刘强
谢大斌
翁志
Original Assignee
北京京东尚科信息技术有限公司
北京京东世纪贸易有限公司
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
Priority claimed from CN201820294180.2U external-priority patent/CN208433046U/zh
Priority claimed from CN201810175168.4A external-priority patent/CN108460908A/zh
Application filed by 北京京东尚科信息技术有限公司, 北京京东世纪贸易有限公司 filed Critical 北京京东尚科信息技术有限公司
Publication of WO2019165895A1 publication Critical patent/WO2019165895A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G07—CHECKING-DEVICES
    • G07F—COIN-FREED OR LIKE APPARATUS
    • G07F9/00—Details other than those peculiar to special kinds or types of apparatus
    • G07F9/009—User recognition or proximity detection
    • G—PHYSICS
    • G07—CHECKING-DEVICES
    • G07F—COIN-FREED OR LIKE APPARATUS
    • G07F11/00—Coin-freed apparatus for dispensing, or the like, discrete articles
    • G07F11/004—Restocking arrangements therefor
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q20/00—Payment architectures, schemes or protocols
    • G06Q20/08—Payment architectures
    • G06Q20/20—Point-of-sale [POS] network systems
    • G06Q20/203—Inventory monitoring
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q20/00—Payment architectures, schemes or protocols
    • G06Q20/08—Payment architectures
    • G06Q20/20—Point-of-sale [POS] network systems
    • G06Q20/208—Input by product or record sensing, e.g. weighing or scanner processing
    • G—PHYSICS
    • G07—CHECKING-DEVICES
    • G07F—COIN-FREED OR LIKE APPARATUS
    • G07F9/00—Details other than those peculiar to special kinds or types of apparatus
    • G07F9/02—Devices for alarm or indication, e.g. when empty; Advertising arrangements in coin-freed apparatus
    • G07F9/026—Devices for alarm or indication, e.g. when empty; Advertising arrangements in coin-freed apparatus for alarm, monitoring and auditing in vending machines or means for indication, e.g. when empty

Definitions

  • the present disclosure relates to the field of vending machines, and more particularly to a vending method and system, as well as vending machines and vending machines.
  • the vending machine purchase process is that the user first discriminates the merchandise through the glass of the vending machine, and then selects the merchandise through a touch screen or a button, and then pays the purchase price through a coin or payment application, and finally the merchandise is The vending machine's shipping port slips out.
  • a vending method comprising:
  • the information of the goods to be purchased by the user is determined according to the weight change information of the shelves;
  • the information of the goods to be purchased by the user is determined according to the video surveillance image in the vending machine
  • the goods information to be purchased by the user is automatically settled.
  • determining whether the vending machine is placed on the shelf is a preset item comprising:
  • determining whether the vending machine is placed on the shelf is a preset item comprising:
  • determining the item information to be purchased by the user according to the weight change information of the shelf comprises:
  • the quantity of the goods purchased by the user and the quantity of the goods are determined based on the weight change information of the shelf in which the weight change occurs, in combination with the weight of the goods carried by the shelf and the weight of the individual goods in which the weight change occurs.
  • determining, according to the video surveillance image in the vending machine, the item information that the user wants to purchase includes:
  • the item information corresponding to the image of the item to be purchased by the user is identified according to a pre-established product feature library.
  • locating the image of the item to be purchased by the user comprises: inputting a video surveillance image in the vending machine into a product positioning model, the item positioning model outputting an image of the item to be purchased by the user
  • the goods positioning model is trained by the image of a plurality of goods and the labeling information of the position of the goods.
  • identifying the item information corresponding to the image of the item to be purchased by the user comprises: inputting an image of the item to be purchased by the user into a product identification model, the item identification model outputting the item to be purchased by the user And corresponding to the product feature, the product identification model is obtained by training the image of the plurality of goods and the tag information of the product classification, and performing the product feature output by the product identification model and the product feature in the product feature database. Matching, and matching the tag information of the product classification corresponding to the product feature in the matched product feature library as the product information corresponding to the image of the product to be purchased by the user.
  • the method further comprises: prior to detecting that the door of the vending machine is closed, reducing weight information and increased weight information based on the weight sensor sensed by the gravity sensor and initial setting of the shelf when the replenishment is completed.
  • the weight of the individual goods is determined, and it is determined whether the user puts the goods to be purchased back to the original shelf, and prompts when the original shelf is not returned, and the returned goods are not settled after being returned to the original shelf.
  • the method further comprises: detecting, according to the video surveillance image in the vending machine, whether the merchandise placement on the shelf of the vending machine is a rule, and issuing a prompt when irregular.
  • the method further comprises:
  • the quantity of the supplementary goods is automatically recorded or reported.
  • the method further comprises:
  • the information is recommended for the user according to the shopping behavior information corresponding to the identity information of the user;
  • At least one of the gender and age of the user is identified according to a face recognition technology, and the user is performed according to at least one of the identified gender and age of the user.
  • Information recommendation is provided.
  • a vending apparatus comprising: a module that performs the vending method of any of the preceding embodiments.
  • a vending apparatus comprising:
  • a processor coupled to the memory, the processor being configured to perform the vending method of any of the preceding embodiments based on instructions stored in the memory.
  • a computer readable storage medium having stored thereon a computer program that, when executed by a processor, implements the vending method of any of the foregoing embodiments.
  • an automatic vending machine comprising:
  • a gravity sensor disposed on a lower surface of each shelf of the vending machine
  • a built-in camera is disposed on the side inner wall of the vending machine corresponding to each shelf, and the shooting direction is aligned with the shelf area;
  • An electromagnetic lock disposed at a cabinet door of the vending machine, capable of opening or closing the cabinet door;
  • the industrial computer is electrically connected to the gravity sensor, the built-in camera, and the electromagnetic lock respectively, and the industrial computer issues an instruction to the electromagnetic lock to open the cabinet when the user identity is effectively recognized.
  • a door and transmitting the weight change information of the shelf perceived by the gravity sensor and the video surveillance image captured by the built-in camera to a vending device in the background, for the vending device to determine the item to be purchased by the user And transmitting information of the door closing to the vending device to trigger the vending device to automatically settle the goods information that the user wants to purchase.
  • the built-in camera is staggered on both side inner walls of the vending machine corresponding to two adjacent shelves.
  • the gravity sensor comprises a gravity sensitive element; or the gravity sensor comprises a gravity sensitive element and a digital conversion element, the gravity sensitive element being electrically connected to the digital conversion element;
  • the gravity sensing component is configured to sense weight information of the shelf carrying goods
  • the digital conversion element is configured to convert weight information of the goods into quantity information of the goods.
  • the vending machine further includes an external camera for capturing a facial image of the user and transmitting to the vending device for the vending device to identify the identity of the user.
  • the vending machine further includes: a touch screen for displaying the content defined by the vending device, or for receiving and presenting information recommended by the vending device in the background to the user .
  • the vending machine further includes: a touch screen for displaying the content defined by the vending device, wherein the content displayed by the touch screen includes the vending device according to the identified user The information recommended by the shopping behavior information corresponding to the identity information, or the information recommended by the vending apparatus based on at least one of the identified gender and age of the user.
  • a vending system comprising: the vending machine of any of the preceding embodiments and the vending apparatus of any of the foregoing embodiments.
  • FIG. 1 is a schematic diagram of some embodiments of a vending system of the present disclosure.
  • FIGS. 2 to 4 are schematic structural views of some embodiments of the vending machine of the present disclosure.
  • FIG. 5 is a schematic flow chart of some embodiments of a vending method according to the present disclosure.
  • FIG. 6 is a schematic flow chart of some embodiments of a vending method of the present disclosure.
  • FIG. 7 is a schematic flow chart of some embodiments of a method for placing goods according to the vending method of the present disclosure.
  • FIG. 8 is a schematic flow chart of some embodiments of a method for replenishing goods in a vending machine related to a vending method according to the present disclosure.
  • FIG. 9 is a schematic flow chart of some embodiments of an information recommendation method related to a vending method according to the present disclosure.
  • Figure 10 is a schematic view showing the structure of some embodiments of the vending apparatus of the present disclosure.
  • Figure 11 is a schematic view showing the structure of some embodiments of the vending apparatus of the present disclosure.
  • the inventor has found that the related art mentioned in the background section has limited information on the products that the user can access before the purchase, and the payment is made first, and if the purchase is wrong, the product cannot be returned, and the shopping experience needs to be improved.
  • One technical problem to be solved by the embodiments of the present disclosure is to enhance the shopping experience under the vending scene.
  • FIG. 1 is a schematic diagram of some embodiments of a vending system of the present disclosure.
  • the vending system includes a vending machine 10 and a vending apparatus 20.
  • the vending machine 10 is in signal communication with the vending apparatus 20.
  • vending device 20 may be, for example, a server or a cloud server.
  • the server or cloud server provides services to a number of vending machines 10.
  • the vending device 20 may also be, for example, a component of the vending machine 10.
  • the vending machine 10 allows the user to open the door of the vending machine 10 and purchase the items in the vending machine 10 in an open environment so that the user can contact and fully understand before purchasing. Goods.
  • the vending apparatus 20 determines the item information to be purchased by the user based on the weight change information of the shelf perceived by the gravity sensor transmitted by the vending machine 10 and the video surveillance image captured by the built-in camera, and receives the delivery information from the vending machine 10. After the information of the door closing is closed, the goods information to be purchased by the user is automatically settled, so that the settlement is simpler, and there is no need to queue or scan the QR code and other additional settlement operations to enhance the shopping experience under the vending scene.
  • FIGS. 2 to 4 are schematic structural views of some embodiments of the vending machine of the present disclosure.
  • the vending machine 10 typically has a multi-layered shelf 11 on which items can be placed.
  • the vending machine 10 is provided with components such as the industrial computer 12, the gravity sensor 13, the built-in camera 14, and the electromagnetic lock 15.
  • the exterior vending machine 10 is provided with components such as an external camera 16 and a touch panel 17.
  • the industrial computer 12 is electrically connected to the gravity sensor 13, the built-in camera 14, the electromagnetic lock 15, the external camera 16, and the touch panel 17, respectively.
  • the industrial computer 12 as a control component can generally be hidden from being disposed inside the vending machine 10.
  • the industrial computer 12 is used to control components of the vending machine 10, such as the electromagnetic lock 15, and may interact with the vending device 20 in the background.
  • the industrial computer 12 can receive the unlocking command issued by the vending apparatus 20 when the user identity is effectively recognized, and then issue an instruction to the electromagnetic lock 15 to open the cabinet door; the industrial computer 12 can also sense the shelf of the gravity sensor 13
  • the weight change information and the video surveillance image captured by the built-in camera 14 are transmitted to the vending apparatus 20 for the vending apparatus 20 to determine the item information that the user wants to purchase; the industrial computer 12 can also transmit the information of the door closing to the vending.
  • the goodsling device 20 automatically triggers the settlement of the goods information that the user wants to purchase by the trigger vending apparatus 20.
  • the gravity sensor 13 can be disposed on the lower surface of each shelf 11 of the vending machine to sense the weight of the item placed on the shelf 11.
  • the gravity sensor 13 can include a gravity sensitive element and a digital conversion element that is electrically coupled to the digital conversion element. Gravity sensitive components are used to sense the weight information of the shelves carrying goods. The digital conversion element is used to convert the weight information of the goods into the quantity information of the goods, and realize the counting function.
  • the gravity sensor 13 may also include only gravity sensitive elements.
  • the built-in camera 14 can be disposed on the side inner wall of the vending machine corresponding to each shelf 11, and the shooting direction is aligned with the shelf area, so that the goods on the shelf 11 and the take-up and return of the goods can be photographed. Case.
  • the built-in camera 14 can be staggered on the inner walls of both sides of the vending machine 10 corresponding to the adjacent two shelves 11. For example, from top to bottom, a built-in camera 14 is disposed on the right inner wall of the vending machine 10 corresponding to the first shelf, and a built-in camera 14 is disposed on the left inner wall of the vending machine 10 corresponding to the second shelf.
  • a built-in camera 14 is disposed on the right inner wall of the vending machine 10 corresponding to the third shelf, and a built-in camera 14 is disposed on the left inner wall of the vending machine 10 corresponding to the fourth shelf.
  • the electromagnetic lock 15 can be placed at the door of the vending machine 10 to open or close the door.
  • an external camera 16 may be disposed on an outer side surface of the cabinet door of the vending machine 10 for taking a facial image of the user and transmitting it to the vending apparatus 20 for the vending apparatus 20 identifies the identity of the user.
  • the touch screen 17, may be disposed on the outside surface of the cabinet door of the vending machine 10 for presenting the content defined by the vending apparatus 20.
  • the content displayed on the touch screen includes: pre-defined recommendation information, such as an advertisement.
  • the touch screen 17 is also used to receive and display information recommended by the back-end vending device to the user of the current purchase.
  • the vending apparatus 20 performs information recommended based on the shopping behavior information corresponding to the identified identity information of the user, or the vending apparatus 20 implements information based on at least one of the identified gender and age of the identified user. Accurate delivery.
  • FIG. 5 is a schematic flow chart of some embodiments of a vending method according to the present disclosure.
  • the vending method of this embodiment includes steps 510-540.
  • step 510 the user selects to open the door of the vending machine by scanning the two-dimensional code or face recognition, and the vending apparatus 20 detects that the user opens the door of the vending machine 10 and opens the door. Obtain the identity information of the user during the process.
  • the external camera 16 of the vending machine 10 detects a human face, and after detecting a human face, the shopping process can be displayed in the touch screen 17.
  • the user is prompted to select the door opening mode, the QR code scanning door opening or the face recognition opening door. If the user selects the QR code to scan the door, the touch screen 17 displays the two-dimensional code, the two-dimensional code carries the device information of the vending machine 10, the user scans the two-dimensional code, and the vending device 20 attempts to acquire the individual representing the user identity. If the personal identification number (PIN) is not obtained, the unlocking fails.
  • PIN personal identification number
  • the vending apparatus 20 issues an unlocking command to the industrial computer 12, and the industrial computer 12 sends the electromagnetic lock 15 to the electromagnetic lock 15. Command to open the door. If the user selects face recognition to open the door, the external camera 16 captures the facial image of the user and transmits it to the vending apparatus 20, and the vending apparatus 20 searches for the facial image feature matching the transmitted facial image feature from the facial image feature library. If the matching facial image feature is not found, the unlocking fails, and if the matching facial image feature is found, the vending apparatus 20 acquires the user identity information corresponding to the matched facial image feature recorded in the feature library, and issues an unlocking instruction to the The industrial computer 12 and the industrial computer 12 issue an instruction to the electromagnetic lock 15 to open the cabinet door.
  • the vending apparatus 20 determines whether the preset item is placed on the shelf of the vending machine to determine whether or not the item is randomly placed.
  • the goods are not placed on the shelves specified in the vending machine, it is considered that the goods are in a random state.
  • the vending apparatus 20 acquires an image of the product currently placed in the vending machine photographed by the built-in camera 14 and completes the image of the current item in the vending machine and the replenishment in the vending machine.
  • the image of the initial placed product is compared, if the item on the currently placed shelf does not match the type of the initially placed item on the shelf, it is determined that the vending machine has non-preset items on the shelf. , indicating the occurrence of goods indiscriminately. On the other hand, if the two are consistent, it means that the preset goods are placed on the shelves, and there is no such thing as random goods.
  • the vending apparatus 20 obtains the weight information of the goods perceived by the gravity sensor 13, and the item weight information may be the total weight or the total quantity of all the items on the shelf, which will be on the shelf currently placed in the vending machine.
  • the weight of the individual item is compared with the weight of the individual item on the shelf initially placed at the completion of the replenishment of the vending machine. If the weights are inconsistent, it is determined that the vending machine has a non-preset on the shelf.
  • the goods indicate the occurrence of the goods in a random manner. On the other hand, if the two are consistent, it means that the preset goods are placed on the shelves, and there is no such thing as random shipments.
  • step 530a in the case where the preset item is placed on the shelf of the vending machine, the vending apparatus 20 determines the item information to be purchased by the user based on the weight change information of the shelf. This method is more efficient in identifying goods information.
  • the shelf and the weight change information of the weight change are determined, and the weight change information of the shelf according to the weight change is combined with the weight change.
  • the gravity sensor S001 is placed under the shelf J001, and the shelf J001 carries the product W001.
  • the gravity sensor S001 senses that the load of the shelf J001 changes, the weight information currently detected by the shelf J001 is reported. If the gravity sensor S001 has a digital conversion function, the number of goods currently carried by the shelf J001 is directly reported, if the gravity sensor S001 If there is no digital conversion function, the weight currently carried by the shelf J001 is reported. Assuming that the quantity of the goods reported by the gravity sensor S001 is 4, the vending apparatus 20 knows that the gravity sensor S001 is placed under the shelf J001, and the shelf J001 carries the goods W001.
  • the vending apparatus 20 can determine that the user wants to purchase the item W001, and the shelf The number of the J001 before the load change is assumed to be 5, and the number after the load change is 4, it can be determined that the number of users who want to purchase the goods W001 is 1.
  • step 530b in the case where non-preset items are placed on the shelf of the vending machine, the vending apparatus 20 determines the item information to be purchased by the user based on the video surveillance image in the vending machine. In this way, the accuracy of identification of goods information is high.
  • the vending apparatus 20 locates an image of the item to be purchased by the user from the video surveillance image in the vending machine, and identifies the item corresponding to the image of the item to be purchased by the user based on the pre-established item feature library. information.
  • Locating the image of the product to be purchased by the user includes: inputting the video surveillance image in the vending machine into the product positioning model, the product positioning model outputting an image of the product to be purchased by the user, the product positioning model passing the image of the plurality of goods and the position of the product
  • the annotation information is obtained by training.
  • the product positioning model can be, for example, a convolutional neural network model.
  • the product positioning model can be trained before use. The product positioning model is trained by the image of multiple items and the labeling information of the position of the goods.
  • an image of a plurality of goods for example, a manual labeling method, marks the position of the goods in the image, and then inputs the image of the plurality of goods and the position information of the goods position into the product positioning model, and the product positioning model
  • the product positioning model Through training, whether the goods and the approximate position of the goods exist in the image are obtained, and the approximate position of the goods identified by the model is compared with the marked goods position, and it is judged whether the comparison result satisfies the requirement of the loss function of the constructed goods position, repeated iteration, optimization and The parameters of the goods positioning model are adjusted so that the comparison result finally satisfies the requirement of the loss function of the constructed goods position, and the product positioning model is saved.
  • the item information corresponding to the image of the item to be purchased by the user includes: inputting an image of the item to be purchased by the user into the item identification model, and the item identification model outputs the item feature corresponding to the image of the item to be purchased by the user, and the item identification model passes the plurality of items.
  • the image and the labeling information of the product classification are trained to match the product features outputted by the product identification model with the product features in the product feature database, and the labeling of the product classification corresponding to the product features in the matched product feature database
  • the information is the item information corresponding to the image of the item to be purchased by the user.
  • the item identification model can be, for example, a convolutional neural network model.
  • the item identification model can be trained prior to use.
  • the product identification model is trained by the image of multiple items and the labeling information of the goods classification. Specifically, the image of the plurality of goods, for example, by manual labeling, labels the classification of the goods in the image, and the classification of the goods can be specific to the SKU (Stock Keeping Unit) level of the goods, not only the goods belonging to the goods.
  • SKU Stock Keeping Unit
  • the category of the goods and the classification information of the goods are input into the product identification model.
  • the product identification model obtains the probability information of the goods in the image by training, and the product with the highest probability can be classified as the product identification.
  • the product feature library can be established by inputting an image of an existing product into a product identification model, and the product identification model outputs a product feature corresponding to the image of the product, wherein the image of the existing product has the labeling information of the product classification, and then, The product feature corresponding to the image of the product output by the product identification model and the tag information of the product classification are stored in the product feature database.
  • step 540 the vending apparatus 20 has acquired the identity information of the user during the opening process of the door. After detecting that the door of the vending machine is closed, the goods information to be purchased by the user may be automatically settled.
  • the user usually closes the door after the purchase is completed, and the electromagnetic lock 15 closes the door.
  • the industrial computer 12 detects that the electromagnetic lock 15 closes the door, the information about the closing of the door is transmitted to the vending.
  • the device 20 triggers a settlement process.
  • the industrial computer 12 can automatically move to the electromagnetic lock if the user does not have any operations such as taking, returning, closing, etc. after detecting the preset time. 15 issues an instruction to close the door, and then transmits the information that the door is closed to the vending apparatus 20, triggering the settlement process.
  • the vending scheme of the above embodiment allows the user to get the goods first and then pay, so that the user can contact and fully understand the goods before the purchase, and automatically settle, the settlement is simpler, and there is no need to queue or scan the QR code and other additional settlement operations. Improve the shopping experience in the vending scene. In addition, regardless of whether the goods in the vending machine are randomly placed, the information of the goods that the user wants to purchase can be accurately identified, and the accuracy and efficiency of the identification of the goods information can be taken into consideration, and the utility model has strong practicability.
  • FIG. 6 is a schematic flow chart of some embodiments of a vending method of the present disclosure.
  • the vending method of this embodiment includes steps 610-650.
  • the user opens the door of the vending machine and begins the self-service shopping process. For example, go through steps 510-530a/530b.
  • the vending apparatus prior to detecting the closing of the door of the vending machine, reduces the weight information and increased weight information based on the shelf perceived by the gravity sensor and the individual items initially placed on the shelf when the replenishment is completed.
  • the weight determines whether the user will return the item to be purchased back to the original shelf.
  • a prompt is issued to remind the user to put the goods back to the original shelf, and the goods that are put back to the shelf but not returned to the original shelf are not settled (step 630), after the original shelf has been put back, Goods returned to the original shelf are not settled (step 640).
  • the weight information therein may be the weight or the number of items that can reflect the weight.
  • a shelf has an increase in weight, it is determined that the user has the behavior of putting back the goods; if a shelf only has an increase in weight, there is no weight reduction. If it is determined that the user has not put the taken goods back to the original shelf, for example, the shelf J001 only has an increase in weight, and without weight reduction, it is determined that the user puts the goods of other shelves on the shelf J001; if the weight of the same shelf occurs In both cases of reduction and weight increase, and the reduced weight information of the shelf is the same as the added weight information, it is determined that the user takes all the goods taken back to the original shelf, for example, the weight information reduced by the shelf J002 is the same as the added weight information.
  • the user takes all the goods taken from the shelf J002 back to the shelf J002; if the same shelf has two cases of weight reduction and weight increase, and the reduced weight information and the added weight information of the shelf are the shelf Integer multiple of the weight of a single item initially placed at the time of replenishment Put the taken goods back to the original shelf.
  • the weight of a single item initially placed on the shelf J003 at the time of replenishment is 100 grams.
  • the weight of the shelf is 200 grams, and the added weight is 100 grams. Two items were taken from the shelf J003, and then one item was returned to the shelf J003.
  • the vending apparatus 20 can automatically settle the item of the user that has not been returned to the shelf.
  • the vending scheme of the above embodiment on the basis of the advantages of the embodiment shown in FIG. 5, can also allow and recognize the behavior of the user to put back the goods, and automatically settle only the goods that are not put back on the shelf, thereby improving the automatic settlement.
  • FIG. 7 is a schematic flow chart of some embodiments of a method for placing goods according to the vending method of the present disclosure.
  • the vending method of this embodiment includes steps 710-730.
  • the built-in camera 14 captures a video surveillance image in the vending machine and transmits it to the vending apparatus 20 via the industrial computer 12.
  • the vending apparatus 20 detects whether the goods on the shelf of the vending machine are placed according to the video surveillance image in the vending machine, and prompts the tally to place the goods in an irregular manner. Rules (step 730).
  • the vending apparatus 20 compares the image of the item currently placed in the vending machine with the image of the item initially placed upon completion of the replenishment in the vending machine, if currently placed on the shelf If the difference between the image of a single item and the image of a single item on the shelf initially placed exceeds the preset range, it is determined that the item on the shelf is irregularly placed, for example, the front side of the item may not be exposed, the product may be tilted or dumped, or even It is the case that the goods are not placed on the specified shelves.
  • the above embodiment proposes a self-service vending scheme with an automatic tally function.
  • FIG. 8 is a schematic flow chart of some embodiments of a method for replenishing goods in a vending machine related to a vending method according to the present disclosure.
  • the vending method of this embodiment includes steps 810-830.
  • step 810 the operation of the refiller opening the cabinet door of the vending machine 10 is detected.
  • the replenisher and the regular user can be distinguished by different identity information.
  • the quantity of the supplemental item is automatically determined based on the weight change information of the shelf sensed by the gravity sensor 13 and the weight of the individual item placed on the shelf.
  • the goods placed on the shelf are not changed, and the replenisher places the goods on the original shelf, and the number of replenished items can be automatically determined according to the weight of the individual goods placed corresponding to the previously recorded shelves. .
  • the goods corresponding to the shelves are changed.
  • the replenisher takes out the original goods on the shelves and places other goods. According to the weight of the other goods, the quantity of the supplementary goods can be automatically determined. .
  • step 830 after detecting that the door of the vending machine 10 is closed, the number of supplementary items is automatically recorded or reported.
  • the amount of the supplemental item can be automatically determined by the gravity sensor 13 and reported to the vending apparatus 20. If the gravity sensor 13 does not have a digital conversion function, the gravity sensor 13 can report the weight change information of the perceived shelf to the vending apparatus 20, and the vending apparatus 20 changes the information according to the weight of the shelf and the individual items of the shelf correspondingly placed. Weight, automatically determines the quantity of supplementary goods. In both ways, the replenisher does not need to manually report the replenishment quantity.
  • the above embodiment proposes a self-service vending program with a self-service replenishment function, which can automatically obtain the replenishment quantity information.
  • FIG. 9 is a schematic flow chart of some embodiments of an information recommendation method related to a vending method according to the present disclosure.
  • the vending method of this embodiment includes steps 910-940.
  • the vending machine 10 plays the recommended information, such as an advertisement, predefined by the vending apparatus 20 through the touch screen 17.
  • the user selects to open the door of the vending machine by scanning the two-dimensional code or face recognition, and the vending apparatus 20 detects that the user opens the door of the vending machine 10 and opens the door. Obtain the identity information of the user during the process.
  • step 930 if the identity information of the user is identified, the information recommendation is performed for the user according to the shopping behavior information corresponding to the identity information of the user, and the recommendation information is played through the touch screen 17.
  • step 940 if the identity information of the user is not recognized, at least one of the gender and the age of the user is recognized according to the face recognition technology, and the user is imaged according to at least one of the gender and age of the identified user.
  • the information recommendation is performed, and the recommendation information is played through the touch screen 17.
  • the above embodiment proposes a self-service vending solution with accurate information push function, which can push information that the user may pay attention to according to the user's shopping behavior, gender and age, so that the information is more accurately placed.
  • Some embodiments of the present disclosure also provide a vending apparatus comprising: a module that performs the vending method of any of the preceding embodiments.
  • Figure 10 is a schematic view showing the structure of some embodiments of the vending apparatus of the present disclosure.
  • the vending apparatus includes:
  • the detecting module 1010 is configured to detect an operation of the user opening the cabinet door of the vending machine
  • the determining module 1020 is configured to determine whether the preset item is placed on the shelf of the vending machine
  • the determining module 1030 is configured to determine, when the preset item is placed on the shelf of the vending machine, the information of the item to be purchased by the user according to the weight change information of the shelf; and the shelf of the vending machine is placed In the case of non-scheduled goods, the information of the goods to be purchased by the user is determined according to the video surveillance image in the vending machine;
  • the settlement module 1040 is configured to automatically check the goods information that the user wants to purchase after the door of the vending machine is closed.
  • the determining module 1020 is configured to compare the image of the product currently placed in the vending machine with the image of the product initially placed when the replenishment machine is completed, if the shelf is currently placed on the shelf. If the goods are inconsistent with the type of goods initially placed on the shelf, it is determined that the vending machine has non-preset items on the shelf.
  • the determining module 1020 is configured to collect the weight of the individual items on the shelf currently placed in the vending machine and the individual items on the shelf that are initially placed when the replenishment machine is completed. The weights are compared, and if the weights are inconsistent, it is determined that the shelves of the vending machine are placed with non-preset items.
  • the determining module 1030 includes a first determining unit, and determining, according to the weight change information of the shelf, the item information that the user wants to purchase includes:
  • the quantity of the goods purchased by the user and the quantity of the goods are determined based on the weight change information of the shelf in which the weight change occurs, in combination with the weight of the goods carried by the shelf and the weight of the individual goods in which the weight change occurs.
  • the determining module 1030 includes a second determining unit, configured to determine, according to the video surveillance image in the vending machine, the item information that the user wants to purchase includes:
  • the item information corresponding to the image of the item to be purchased by the user is identified according to the pre-established product feature library.
  • locating the image of the item to be purchased by the user comprises: inputting a video surveillance image in the vending machine into the item positioning model, the item positioning model outputting an image of the item to be purchased by the user, and the item positioning model passes the plurality of items The image and the labeling information of the position of the goods are trained.
  • identifying the item information corresponding to the image of the item to be purchased by the user comprises: inputting an image of the item to be purchased by the user into the item identification model, and the item identification model outputs the item feature corresponding to the image of the item to be purchased by the user, the item
  • the recognition model is obtained by training the images of the plurality of goods and the labeling information of the goods classification, matching the product features outputted by the product identification model with the product features in the product feature database, and matching the product features in the product feature database.
  • the label information of the corresponding item classification is the item information corresponding to the image of the item to be purchased by the user.
  • the determining module 1020 is further configured to: before detecting the closing of the door of the vending machine, the weight information and the increased weight information of the shelf sensed by the gravity sensor and the initial pendulum of the shelf when the replenishment is completed.
  • the weight of a single item is determined to determine whether the user returns the item to be purchased back to the original shelf, and when the original shelf is not returned, a prompt is issued, and after the original shelf has been placed back, the returned item is not settled.
  • the method further includes: a tally module 1050, configured to detect, according to the video surveillance image in the vending machine, whether the merchandise on the shelf of the vending machine is placed or not, and prompts when irregular.
  • a tally module 1050 configured to detect, according to the video surveillance image in the vending machine, whether the merchandise on the shelf of the vending machine is placed or not, and prompts when irregular.
  • the method further includes: a replenishment module 1060, configured to detect an operation of the replenisher opening the cabinet door of the vending machine; a weight change information of the shelf sensed by the gravity sensor; and a single item corresponding to the shelf The weight determines the quantity of the supplementary goods; after detecting that the door of the vending machine is closed, the quantity of the supplementary goods is automatically recorded or reported.
  • a replenishment module 1060 configured to detect an operation of the replenisher opening the cabinet door of the vending machine; a weight change information of the shelf sensed by the gravity sensor; and a single item corresponding to the shelf The weight determines the quantity of the supplementary goods; after detecting that the door of the vending machine is closed, the quantity of the supplementary goods is automatically recorded or reported.
  • the information recommendation module 1070 is configured to: if the identity information of the user is identified, perform information recommendation for the user according to the shopping behavior information corresponding to the identity information of the user; if the identity information of the user is not recognized, according to The face recognition technology recognizes at least one of the gender and age of the user, and performs information recommendation for the user according to at least one of the gender and age of the identified user.
  • Figure 11 is a schematic view showing the structure of some embodiments of the vending apparatus of the present disclosure.
  • the apparatus 1100 of this embodiment includes a memory 1110 and a processor 1120 coupled to the memory 1110, the processor 1120 being configured to perform any of the foregoing embodiments based on instructions stored in the memory 1110.
  • the vending method includes a memory 1110 and a processor 1120 coupled to the memory 1110, the processor 1120 being configured to perform any of the foregoing embodiments based on instructions stored in the memory 1110. The vending method.
  • the memory 1110 may include, for example, a system memory, a fixed non-volatile storage medium, or the like.
  • the system memory stores, for example, an operating system, an application, a boot loader, and other programs.
  • the device 1100 can also include an input and output interface 1130, a network interface 1140, a storage interface 1150, and the like. These interfaces 1130, 1140, 1150 and the memory 1110 and the processor 1120 can be connected, for example, via a bus 1160.
  • the input/output interface 1130 provides a connection interface for input and output devices such as a display, a mouse, a keyboard, and a touch screen.
  • Network interface 1140 provides a connection interface for various networked devices.
  • the storage interface 1150 provides a connection interface for an external storage device such as an SD card or a USB flash drive.
  • the present disclosure also proposes a computer readable storage medium having stored thereon a computer program that, when executed by a processor, implements the vending method of any of the foregoing embodiments.
  • embodiments of the present disclosure can be provided as a method, system, or computer program product. Accordingly, the present disclosure may take the form of an entirely hardware embodiment, an entirely software embodiment, or a combination of software and hardware aspects. Moreover, the present disclosure may take the form of a computer program product embodied on one or more computer-usable non-transitory storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer usable program code. .
  • the computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture comprising the instruction device.
  • the apparatus implements the functions specified in one or more blocks of a flow or a flow and/or block diagram of the flowchart.
  • These computer program instructions can also be loaded onto a computer or other programmable data processing device such that a series of operational steps are performed on a computer or other programmable device to produce computer-implemented processing for execution on a computer or other programmable device.
  • the instructions provide steps for implementing the functions specified in one or more of the flow or in a block or blocks of a flow diagram.

Landscapes

  • Business, Economics & Management (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Strategic Management (AREA)
  • General Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Control Of Vending Devices And Auxiliary Devices For Vending Devices (AREA)

Abstract

一种自动售货方法和系统以及自动售货装置(20)和自动贩售机(10)，自动售货方法包括：检测到用户打开自动贩售机(10)的柜门的操作(510)；判断货架(11)上摆放的是否为预设货品(520)；如果是，根据货架(11)的重量变化信息确定用户欲购买的货品信息(530a)；如果否，根据自动贩售机(10)内的视频监控图像确定用户欲购买的货品信息(530b)；检测到柜门关闭后，自动对用户欲购买的货品信息进行结算(540)。自动售货方法和系统以及自动售货装置(20)和自动贩售机(10)允许用户先拿到货品再付款，用户在购买前可接触并充分了解货品，且自动结算，结算更加简单，提升自动售货场景下的购物体验。此外，不论自动贩售机内(10)货品是否乱放，均能够准确识别用户欲购买的货品信息，同时还可以兼顾货品信息识别的准确性和效率，具有较强的实用性。

Description

自动售货方法和系统以及自动售货装置和自动贩售机
交叉引用
本申请是以CN申请号为201810175168.4，申请日为2018年3月2日的申请，和CN申请号为201820294180.2，申请日为2018年3月2日的申请为基础，并主张其优先权，这些CN申请的公开内容在此作为整体引入本申请中。
技术领域
本公开涉及自动贩售机领域，特别涉及一种自动售货方法和系统以及自动售货装置和自动贩售机。
背景技术
随着技术的发展，无人超市、无人便利店成为颠覆传统零售的变革形态，各种自动售货机也应运而生。
相关技术中，自动售货机的购买流程是，首先用户透过自动售货机的玻璃来辨别商品，然后通过触屏或按键选定商品，接着通过投币或支付应用等方式支付货款，最后商品从自动售货机的出货口滑出。
发明内容
根据本公开的一个方面，提出一种自动售货方法，包括：
检测到用户打开自动贩售机的柜门的操作；
判断所述自动贩售机的货架上摆放的是否为预设货品；
在所述自动贩售机的货架上摆放的是预设货品的情况下，根据货架的重量变化信息确定所述用户欲购买的货品信息；
在所述自动贩售机的货架上摆放有非预设货品的情况下，根据所述自动贩售机内的视频监控图像确定所述用户欲购买的货品信息；
检测到所述自动贩售机的柜门关闭后，自动对所述用户欲购买的货品信息进行结算。
在一些实施例中，判断所述自动贩售机的货架上摆放的是否为预设货品包括：
将所述自动贩售机内当前摆放的货品图像与所述自动贩售机内补货完成时初始 摆放的货品图像进行比较，如果当前摆放的货架上的货品与初始摆放的该货架上的货品的种类不一致，则判定所述自动贩售机的该货架上摆放有非预设货品。
在一些实施例中，判断所述自动贩售机的货架上摆放的是否为预设货品包括：
将所述自动贩售机内当前摆放的货架上的单个货品的重量与所述自动贩售机内补货完成时初始摆放的该货架上的单个货品的重量进行比较，如果重量不一致，则判定所述自动贩售机的该货架上摆放有非预设货品。
在一些实施例中，根据货架的重量变化信息确定所述用户欲购买的货品信息包括：
根据重力传感器感知的重量变化信息以及所述重力传感器对应的货架信息，确定发生重量变化的货架及其重量变化信息；
根据发生重量变化的货架的重量变化信息，结合发生重量变化的货架承载的货品和单个货品的重量，确定所述用户购买的货品以及货品的数量。
在一些实施例中，根据所述自动贩售机内的视频监控图像确定所述用户欲购买的货品信息包括：
从所述自动贩售机内的视频监控图像中定位所述用户欲购买的货品的图像；
根据预先建立的货品特征库，识别所述用户欲购买的货品的图像对应的货品信息。
在一些实施例中，定位所述用户欲购买的货品的图像包括：将所述自动贩售机内的视频监控图像输入货品定位模型，所述货品定位模型输出所述用户欲购买的货品的图像，所述货品定位模型通过多个货品的图像及其货品位置的标注信息进行训练得到。
在一些实施例中，识别所述用户欲购买的货品的图像对应的货品信息包括：将所述用户欲购买的货品的图像输入货品识别模型，所述货品识别模型输出所述用户欲购买的货品的图像对应的货品特征，所述货品识别模型通过多个货品的图像及其货品分类的标注信息进行训练得到，将所述货品识别模型输出的货品特征与所述货品特征库中的货品特征进行匹配，并将匹配到的所述货品特征库中的货品特征对应的货品分类的标注信息作为所述用户欲购买的货品的图像对应的货品信息。
在一些实施例中，该方法还包括：在检测到所述自动贩售机的柜门关闭前，根据重力传感器感知的货架减少的重量信息和增加的重量信息以及货架在补货完成时初始摆放的单个货品的重量，确定所述用户是否将欲购买的货品放回原始货架，在未 放回原始货架时发出提示，在已放回原始货架后，对该放回的货品不进行结算。
在一些实施例中，该方法还包括：根据所述自动贩售机内的视频监控图像，检测所述自动贩售机的货架上的货品摆放是否规则，在不规则时发出提示。
在一些实施例中，该方法还包括：
检测到补货员打开自动贩售机的柜门的操作；
根据重力传感器感知的货架的重量变化信息以及货架对应摆放的单个货品的重量，确定补充货品的数量；
检测到所述自动贩售机的柜门关闭后，自动记录或上报补充货品的数量。
在一些实施例中，该方法还包括：
如果识别到所述用户的身份信息，根据所述用户的身份信息对应的购物行为信息为所述用户进行信息推荐；
如果未识别到所述用户的身份信息，根据人脸识别技术识别所述用户的性别和年龄中的至少一项，根据识别的所述用户的性别和年龄中的至少一项为所述用户进行信息推荐。
根据本公开的另一个方面，提出一种自动售货装置，包括：执行前述任一个实施例中的自动售货方法的模块。
根据本公开的另一个方面，提出一种自动售货装置，包括：
存储器；以及
耦接至所述存储器的处理器，所述处理器被配置为基于存储在所述存储器中的指令，执行前述任一个实施例中的自动售货方法。
根据本公开的另一个方面，提出一种计算机可读存储介质，其上存储有计算机程序，该程序被处理器执行时实现前述任一个实施例中的自动售货方法。
根据本公开的另一个方面，提出一种自动贩售机，包括：
工控机；
重力传感器，设置于所述自动贩售机的每层货架的下表面；
内置摄像头，设置于每层货架对应的所述自动贩售机的侧面内壁，拍摄方向对准货架区域；
电磁锁，设置于所述自动贩售机的柜门处，能够打开或关闭所述柜门；
其中，所述工控机与所述重力传感器、所述内置摄像头、所述电磁锁分别电连接，所述工控机在用户身份被有效识别的情况下，向所述电磁锁发出指令打开所述柜门， 并将所述重力传感器感知的货架的重量变化信息和所述内置摄像头拍摄的视频监控图像传送至后台的自动售货装置，以供所述自动售货装置确定所述用户欲购买的货品信息，并将所述柜门关闭的信息传送至所述自动售货装置，以触发所述自动售货装置自动对所述用户欲购买的货品信息进行结算。
在一些实施例中，所述内置摄像头在相邻两层货架对应的所述自动贩售机的两侧面内壁交错设置。
在一些实施例中，所述重力传感器包括重力敏感元件；或者，所述重力传感器包括重力敏感元件和数字变换元件，所述重力敏感元件与所述数字变换元件电连接；
所述重力敏感元件用于感知货架承载货品的重量信息；
所述数字变换元件用于将货品的重量信息变换为货品的数量信息。
在一些实施例中，自动贩售机还包括：外置摄像头，用于拍摄用户的面部图像，并传送至所述自动售货装置，以供所述自动售货装置识别所述用户的身份。
在一些实施例中，自动贩售机还包括：触摸屏，用于展示所述自动售货装置定义的内容，或者，用于接收并展示后台的所述自动售货装置向所述用户推荐的信息。
在一些实施例中，自动贩售机还包括：触摸屏，用于展示所述自动售货装置定义的内容，其中，所述触摸屏展示的内容包括所述自动售货装置根据识别的所述用户的身份信息对应的购物行为信息推荐的信息，或者，所述自动售货装置根据识别的所述用户的性别和年龄中的至少一项推荐的信息。
根据本公开的另一个方面，提出一种自动售货系统，包括：前述任一个实施例中的自动贩售机和前述任一个实施例中的自动售货装置。
附图说明
下面将对实施例或相关技术描述中所需要使用的附图作简单地介绍。根据下面参照附图的详细描述，可以更加清楚地理解本公开，
显而易见地，下面描述中的附图仅仅是本公开的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。
图1为本公开自动售货系统一些实施例的示意图。
图2～4为本公开自动售货机一些实施例的结构示意图。
图5为本公开自动售货方法一些实施例的流程示意图。
图6为本公开自动售货方法一些实施例的流程示意图。
图7为本公开与自动售货方法相关的货品摆放方法一些实施例的流程示意图。
图8为本公开与自动售货方法相关的自动贩售机中货品补充方法一些实施例的流程示意图。
图9为本公开与自动售货方法相关的信息推荐方法一些实施例的流程示意图。
图10为本公开自动售货装置一些实施例的结构示意图。
图11为本公开自动售货装置一些实施例的结构示意图。
具体实施方式
下面将结合本公开实施例中的附图，对本公开实施例中的技术方案进行清楚、完整地描述。
发明人发现，背景技术部分提及的相关技术，用户购买之前所能接触到的商品信息有限，并且是先付款再拿到商品，如果购买错误，将无法退回，购物体验有待提升。
本公开实施例所要解决的一个技术问题是：提升自动售货场景下的购物体验。
图1为本公开自动售货系统一些实施例的示意图。
如图1所示，自动售货系统包括自动贩售机10和自动售货装置20。自动贩售机10与自动售货装置20之间信号连通。在一些实施例中，自动售货装置20例如可以服务器或云服务器。服务器或云服务器为若干个自动贩售机10提供服务。在一些实施例中，自动售货装置20例如还可以是自动贩售机10中的一个部件。
在自动售货应用场景中，自动贩售机10允许用户打开自动贩售机10的柜门，在开放环境中选购自动贩售机10内的货品，使得用户在购买前可以接触并充分了解货品。然后自动售货装置20根据自动贩售机10传送的重力传感器感知的货架的重量变化信息和内置摄像头拍摄的视频监控图像确定用户欲购买的货品信息，并在接收到自动贩售机10传送的柜门关闭的信息后，对用户欲购买的货品信息进行自动结算，使得结算更加简单，无需排队或扫描二维码等额外的结算操作，提升自动售货场景下的购物体验。
图2～4为本公开自动售货机一些实施例的结构示意图。
如图2～4所示，自动贩售机10通常具有多层货架11，货架11上可以摆放货品。自动贩售机10内设置了工控机12、重力传感器13、内置摄像头14、电磁锁15等部件。自动贩售机10外设置了外置摄像头16、触摸屏17等部件。工控机12与重力传感器13、内置摄像头14、电磁锁15、外置摄像头16、触摸屏17分别电连接。
在一些实施例中，工控机12作为控制部件，通常可以隐藏设置于自动贩售机10的内部。在一些实施例中，工控机12用来控制自动贩售机10的例如电磁锁15等部件，并可以与后台的自动售货装置20进行信息交互。例如，工控机12可以接收自动售货装置20在用户身份被有效识别的情况下发出的开锁指令，然后向电磁锁15发出指令打开柜门；工控机12还可以将重力传感器13感知的货架的重量变化信息和内置摄像头14拍摄的视频监控图像传送至自动售货装置20，以供自动售货装置20确定用户欲购买的货品信息；工控机12还可以将柜门关闭的信息传送至自动售货装置20，以触发自动售货装置20自动对用户欲购买的货品信息进行结算。
在一些实施例中，重力传感器13可以设置于自动贩售机的每层货架11的下表面，能够感应货架11上摆放货品的重量。在一些实施例中，重力传感器13可以包括重力敏感元件和数字变换元件，重力敏感元件与数字变换元件电连接。重力敏感元件用于感知货架承载货品的重量信息。数字变换元件用于将货品的重量信息变换为货品的数量信息，实现计数功能。在一些实施例中，重力传感器13还可以仅包括重力敏感元件。
在一些实施例中，内置摄像头14可以设置于每层货架11对应的自动贩售机的侧面内壁，拍摄方向对准货架区域，从而能够拍摄到货架11上的货品以及货品的拿取和放回的情况。为了更全面地拍摄货品以及货品取放的情况，内置摄像头14可以在相邻两层货架11对应的自动贩售机10的两侧面内壁交错设置。例如，从上往下，在第一层货架对应的自动贩售机10的右侧内壁上设置内置摄像头14，在第二层货架对应的自动贩售机10的左侧内壁上设置内置摄像头14，在第三层货架对应的自动贩售机10的右侧内壁上设置内置摄像头14，在第四层货架对应的自动贩售机10的左侧内壁上设置内置摄像头14。
在一些实施例中，电磁锁15可以设置于自动贩售机10的柜门处，能够打开或关闭柜门。
在一些实施例中，外置摄像头16，例如可以设置在自动贩售机10的柜门的外侧表面，用于拍摄用户的面部图像，并传送至自动售货装置20，以供自动售货装置20识别用户的身份。
在一些实施例中，触摸屏17，例如可以设置在自动贩售机10的柜门的外侧表面，用于展示自动售货装置20定义的内容。其中，触摸屏展示的内容包括：预先定义的推荐信息，例如广告等。触摸屏17还用于接收并展示后台的自动售货装置向当前购 物的用户推荐的信息。例如，自动售货装置20根据识别的用户的身份信息对应的购物行为信息推荐的信息，或者，自动售货装置20根据识别的用户的性别和年龄中的至少一项推荐的信息，从而实现信息的精准投放。
下面描述自动售货过程。
图5为本公开自动售货方法一些实施例的流程示意图。
如图5所示，该实施例的自动售货方法包括：步骤510-540。
在步骤510，用户选择通过扫描二维码或人脸识别的方式打开自动贩售机的柜门，自动售货装置20检测用户打开自动贩售机10的柜门的操作，并在柜门打开过程中获取用户的身份信息。
在一些实施例中，自动贩售机10的外置摄像头16会检测人脸，在检测到人脸后，可以在触摸屏17中展示购物流程。购物流程中提示用户选择开门方式，二维码扫描开门或人脸识别开门。若用户选择二维码扫描开门，触摸屏17中展示二维码，二维码中带有自动贩售机10的设备信息，用户扫描二维码，自动售货装置20尝试获取代表用户身份的个人识别密码(Personal Identification Number，PIN)，如果未获取到个人识别密码，开锁失败，如果获取到个人识别密码，自动售货装置20下发开锁指令给工控机12，工控机12向电磁锁15发出指令打开柜门。若用户选择人脸识别开门，外置摄像头16拍摄用户的面部图像，并传送至自动售货装置20，自动售货装置20从面部图像特征库查找与传送的面部图像特征匹配的面部图像特征，如果未查找到匹配的面部图像特征，开锁失败，如果查找到匹配的面部图像特征，自动售货装置20获取特征库中记录的匹配的面部图像特征对应的用户身份信息，并下发开锁指令给工控机12，工控机12向电磁锁15发出指令打开柜门。
在步骤520，自动售货装置20判断自动贩售机的货架上摆放的是否为预设货品，以确定是否出现货品乱放的情况。
例如，如果货品未摆放到自动贩售机中规定的货架上，则被认为出现货品乱放的情况。
在一些实施例中，自动售货装置20获取内置摄像头14拍摄的自动贩售机内当前摆放的货品图像，将自动贩售机内当前摆放的货品图像与自动贩售机内补货完成时初始摆放的货品图像进行比较，如果当前摆放的货架上的货品与初始摆放的该货架上的货品的种类不一致，则判定自动贩售机的该货架上摆放有非预设货品，说明出现货品乱放的情况。反之，如果二者一致，则说明货架上摆放的是预设货品，未出现货品乱 放的情况。
在一些实施例中，自动售货装置20获取重力传感器13感知的货品重量信息，该货品重量信息可以是货架上所有货品的总重量或总数量，将自动贩售机内当前摆放的货架上的单个货品的重量与自动贩售机内补货完成时初始摆放的该货架上的单个货品的重量进行比较，如果重量不一致，则判定自动贩售机的该货架上摆放有非预设货品，说明出现货品乱放的情况。反之，如果二者一致，则说明货架上摆放的是预设货品，未出现货品乱放的情况。
在步骤530a，在自动贩售机的货架上摆放的是预设货品的情况下，自动售货装置20根据货架的重量变化信息确定用户欲购买的货品信息。这种方式货品信息识别效率较高。
在一些实施例中，根据重力传感器感知的重量变化信息以及重力传感器对应的货架信息，确定发生重量变化的货架及其重量变化信息，根据发生重量变化的货架的重量变化信息，结合发生重量变化的货架承载的货品和单个货品的重量，确定用户购买的货品以及货品的数量。
例如，重力传感器S001放置在货架J001下，货架J001承载货品W001。重力传感器S001感知到货架J001的承重发生变化时，上报感应到的货架J001当前承载的重量信息，若重力传感器S001具有数字变换功能，则直接上报货架J001当前承载的货品的数量，若重力传感器S001不具有数字变换功能，则上报货架J001当前承载的重量。假设重力传感器S001上报货品数量为4，自动售货装置20知道重力传感器S001放置在货架J001下，货架J001承载货品W001，因此，自动售货装置20可以确定用户欲购买的是货品W001，并且货架J001的承重发生变化之前的数量假设为5，承重发生变化之后的数量为4，则可以确定用户欲购买货品W001的数量是1。
在步骤530b，在自动贩售机的货架上摆放有非预设货品的情况下，自动售货装置20根据自动贩售机内的视频监控图像确定用户欲购买的货品信息。这种方式货品信息识别的准确率较高。
在一些实施例中，自动售货装置20从自动贩售机内的视频监控图像中定位用户欲购买的货品的图像，根据预先建立的货品特征库，识别用户欲购买的货品的图像对应的货品信息。
定位用户欲购买的货品的图像包括：将自动贩售机内的视频监控图像输入货品定位模型，货品定位模型输出用户欲购买的货品的图像，货品定位模型通过多个货品的 图像及其货品位置的标注信息进行训练得到。货品定位模型例如可以是卷积神经网络模型。此外，货品定位模型在使用之前可以先进行训练。货品定位模型通过多个货品的图像及其货品位置的标注信息进行训练得到。具体地说，对多个货品的图像，例如采用人工标注的方法，标注图像中的货品位置，然后，将多个货品的图像及其货品位置的标注信息输入至货品定位模型中，货品定位模型通过训练得到图像中是否存在货品以及货品的大致位置，将模型识别的货品的大致位置与标注的货品位置进行比较，判断比较结果是否满足构建的货品位置的损失函数的要求，反复迭代，优化和调整货品定位模型的参数，使得比较结果最终满足构建的货品位置的损失函数的要求，保存该货品定位模型。
识别用户欲购买的货品的图像对应的货品信息包括：将用户欲购买的货品的图像输入货品识别模型，货品识别模型输出用户欲购买的货品的图像对应的货品特征，货品识别模型通过多个货品的图像及其货品分类的标注信息进行训练得到，将货品识别模型输出的货品特征与货品特征库中的货品特征进行匹配，并将匹配到的货品特征库中的货品特征对应的货品分类的标注信息作为用户欲购买的货品的图像对应的货品信息。货品识别模型例如可以是卷积神经网络模型。此外，货品识别模型在使用之前可以先进行训练。货品识别模型通过多个货品的图像及其货品分类的标注信息进行训练得到。具体地说，对多个货品的图像，例如采用人工标注的方法，标注图像中的货品分类，该货品分类可以具体到货品的SKU(Stock Keeping Unit，库存量单位)级别，而不仅是货品所属的类别；然后，将多个货品的图像及其货品分类的标注信息输入至货品识别模型中，货品识别模型通过训练得到图像中货品分类的概率信息，可以将概率最大的货品分类确定为货品识别结果；然后，将模型识别的货品分类与标注的货品分类进行比较，判断比较结果是否满足构建的货品分类的损失函数的要求，反复迭代，优化和调整货品识别模型的参数，使得比较结果最终满足构建的货品分类的损失函数的要求，保存该货品识别模型。货品特征库可以采用以下方法建立：将已有货品的图像输入到货品识别模型，货品识别模型输出货品的图像对应的货品特征，其中，已有货品的图像具有货品分类的标注信息，然后，将货品识别模型输出的货品的图像对应的货品特征及其货品分类的标注信息一起存储到货品特征库。
在步骤540，自动售货装置20在柜门打开过程中已经获取到用户的身份信息，在检测到自动贩售机的柜门关闭后，可以自动对该用户欲购买的货品信息进行结算。
在一些实施例中，用户在购买完成后通常会关闭柜门，电磁锁15会关闭柜门， 工控机12检测到电磁锁15关闭柜门后，会将柜门关闭的信息传送至自动售货装置20，触发结算流程。在一些实施例中，若用户在购买完成后忘记关闭柜门，工控机12若检测到超过预设时间后用户仍没有任何的例如拿取、放回、关门等操作，则可以自动向电磁锁15发出关闭柜门的指令，然后将柜门关闭的信息传送至自动售货装置20，触发结算流程。
上述实施例的售货方案，允许用户先拿到货品再付款，使得用户在购买前可以接触并充分了解货品，并且自动结算，结算更加简单，无需排队或扫描二维码等额外的结算操作，提升自动售货场景下的购物体验。此外，不论自动贩售机内货品是否乱放，均能够准确识别用户欲购买的货品信息，同时还可以兼顾货品信息识别的准确性和效率，具有较强的实用性。
图6为本公开自动售货方法一些实施例的流程示意图。
如图6所示，该实施例的自动售货方法包括：步骤610-650。
在步骤610，用户打开自动贩售机的柜门，开始自助购物流程。例如，经过步骤510-530a/530b。
在步骤620，自动售货装置20在检测到自动贩售机的柜门关闭前，根据重力传感器感知的货架减少的重量信息和增加的重量信息以及货架在补货完成时初始摆放的单个货品的重量，确定用户是否将欲购买的货品放回原始货架。在未放回原始货架时发出提示，提醒用户将货品放回原始货架，对于放回货架但未放回原始货架的货品也不进行结算(步骤630)，在已放回原始货架后，对该放回原始货架的货品不进行结算(步骤640)。
其中的重量信息可以是重量或能够反映重量的货品数量。
在一些实施例中，在一个用户的一次购物过程中：如果一个货架发生重量增加的情况，则判定用户有放回货品的行为；如果一个货架仅发生重量增加的情况，而没有重量减少的情况，判定用户未将拿取的货品放回原始货架，例如货架J001仅发生重量增加的情况，而没有重量减少的情况，判定用户将其他货架的货品放到了货架J001上；如果同一个货架发生重量减少和重量增加的两种情况，且该货架减少的重量信息和增加的重量信息相同，判定用户将拿取的货品全部放回原始货架，例如货架J002减少的重量信息和增加的重量信息相同，判定用户将从货架J002拿取的货品全部放回到货架J002上；如果同一个货架发生重量减少和重量增加的两种情况，且该货架减少的重量信息和增加的重量信息均是该货架在补货完成时初始摆放的单个货品的 重量的整数倍，判定用户将拿取的货品放回原始货架，例如货架J003在补货完成时初始摆放的单个货品的重量是100克，该货架减少的重量是200克，增加的重量是100克，判定用户可能先从货架J003拿取了2件货品，后来又放回了1件货品到货架J003。
在步骤650，自动售货装置20在检测到自动贩售机的柜门关闭后，可以自动对该用户的未放回货架的货品进行结算。
上述实施例的售货方案，在图5所示实施例的优点的基础上，还能够允许和识别用户放回货品的行为，并仅对未放回货架的货品进行自动结算，提升了自动结算的准确性，提升了用户的购物体验。
图7为本公开与自动售货方法相关的货品摆放方法一些实施例的流程示意图。
如图7所示，该实施例的自动售货方法包括：步骤710-730。
在步骤710，内置摄像头14拍摄自动贩售机内的视频监控图像，通过工控机12传送至自动售货装置20。
在步骤720，自动售货装置20根据自动贩售机内的视频监控图像，检测自动贩售机的货架上的货品摆放是否规则，在不规则时发出提示，提醒理货员将货品摆放规则(步骤730)。
在一些实施例中，自动售货装置20将自动贩售机内当前摆放的货品图像与自动贩售机内补货完成时初始摆放的货品图像进行比较，如果当前摆放的货架上的单个货品图像与初始摆放的该货架上的单个货品图像的差距超过预设范围，则判定该货架上的货品摆放不规则，例如，可能出现了货品正面未外露，货品倾斜或倾倒，甚至是货品未摆放在规定货架上等情况。
上述实施例提出了一种具有自动理货功能的自助售货方案。
图8为本公开与自动售货方法相关的自动贩售机中货品补充方法一些实施例的流程示意图。
如图8所示，该实施例的自动售货方法包括：步骤810-830。
在步骤810，检测到补货员打开自动贩售机10的柜门的操作。
在一些实施例中，可以通过不同的身份信息，区分补货员与普通用户。
在步骤820，根据重力传感器13感知的货架的重量变化信息以及货架对应摆放的单个货品的重量，自动确定补充货品的数量。
在直接补货的应用场景中，货架对应摆放的货品没有发生变化，补货员将货品放置在原货架上，根据之前记录的货架对应摆放的单个货品的重量，可以自动确定补充 货品的数量。
在更换货品的应用场景中，货架对应摆放的货品发了变化，补货员将货架上的原货品取出，并放置其他货品，根据单个的该其他货品的重量，可以自动确定补充货品的数量。
在步骤830，检测到自动贩售机10的柜门关闭后，自动记录或上报补充货品的数量。
在一些实施例中，如果重力传感器13具有数字变换功能，可以由重力传感器13自动确定补充货品的数量，并上报给自动售货装置20。如果重力传感器13不具有数字变换功能，重力传感器13可以将感知的货架的重量变化信息上报给自动售货装置20，自动售货装置20根据货架的重量变化信息以及货架对应摆放的单个货品的重量，自动确定补充货品的数量。两种方式均不需要补货员人工上报补货数量。
上述实施例提出了一种具有自助补货功能的自助售货方案，可以自动获取补货数量信息。
图9为本公开与自动售货方法相关的信息推荐方法一些实施例的流程示意图。
如图9所示，该实施例的自动售货方法包括：步骤910-940。
在步骤910，在没有用户靠近自动贩售机10时，自动贩售机10通过触摸屏17播放自动售货装置20预先定义的推荐信息，例如广告。
在步骤920，用户选择通过扫描二维码或人脸识别的方式打开自动贩售机的柜门，自动售货装置20检测用户打开自动贩售机10的柜门的操作，并在柜门打开过程中获取用户的身份信息。
在步骤930，如果识别到用户的身份信息，根据用户的身份信息对应的购物行为信息为用户进行信息推荐，推荐信息通过触摸屏17播放。
在步骤940，如果未识别到用户的身份信息，根据人脸识别技术识别用户的性别和年龄中的至少一项，对用户进行画像，根据识别的用户的性别和年龄中的至少一项为用户进行信息推荐，推荐信息通过触摸屏17播放。
上述实施例提出了一种具有信息精准推送功能的自助售货方案，可以根据用户的购物行为、性别和年龄等信息相应的推送用户可能关注的信息，使得信息投放地更加精准。
本公开的一些实施例还提出一种自动售货装置，包括：执行前述任一个实施例的自动售货方法的模块。图10为本公开自动售货装置一些实施例的结构示意图。
如图10所示，自动售货装置，包括：
检测模块1010，用于检测到用户打开自动贩售机的柜门的操作；
判断模块1020，用于判断自动贩售机的货架上摆放的是否为预设货品；
确定模块1030，用于在自动贩售机的货架上摆放的是预设货品的情况下，根据货架的重量变化信息确定用户欲购买的货品信息；在自动贩售机的货架上摆放有非预设货品的情况下，根据自动贩售机内的视频监控图像确定用户欲购买的货品信息；
结算模块1040，用于检测到自动贩售机的柜门关闭后，自动对用户欲购买的货品信息进行结算。
在一些实施例中，判断模块1020，用于将自动贩售机内当前摆放的货品图像与自动贩售机内补货完成时初始摆放的货品图像进行比较，如果当前摆放的货架上的货品与初始摆放的该货架上的货品的种类不一致，则判定自动贩售机的该货架上摆放有非预设货品。
在一些实施例中，判断模块1020，用于将自动贩售机内当前摆放的货架上的单个货品的重量与自动贩售机内补货完成时初始摆放的该货架上的单个货品的重量进行比较，如果重量不一致，则判定自动贩售机的该货架上摆放有非预设货品。
在一些实施例中，确定模块1030，包括第一确定单元，用于根据货架的重量变化信息确定用户欲购买的货品信息包括：
根据重力传感器感知的重量变化信息以及重力传感器对应的货架信息，确定发生重量变化的货架及其重量变化信息；
根据发生重量变化的货架的重量变化信息，结合发生重量变化的货架承载的货品和单个货品的重量，确定用户购买的货品以及货品的数量。
在一些实施例中，确定模块1030，包括第二确定单元，用于根据自动贩售机内的视频监控图像确定用户欲购买的货品信息包括：
从自动贩售机内的视频监控图像中定位用户欲购买的货品的图像；
根据预先建立的货品特征库，识别用户欲购买的货品的图像对应的货品信息。
在一些实施例中，定位用户欲购买的货品的图像包括：将自动贩售机内的视频监控图像输入货品定位模型，货品定位模型输出用户欲购买的货品的图像，货品定位模型通过多个货品的图像及其货品位置的标注信息进行训练得到。
在一些实施例中，识别用户欲购买的货品的图像对应的货品信息包括：将用户欲购买的货品的图像输入货品识别模型，货品识别模型输出用户欲购买的货品的图像对 应的货品特征，货品识别模型通过多个货品的图像及其货品分类的标注信息进行训练得到，将货品识别模型输出的货品特征与货品特征库中的货品特征进行匹配，并将匹配到的货品特征库中的货品特征对应的货品分类的标注信息作为用户欲购买的货品的图像对应的货品信息。
在一些实施例中，判断模块1020，还用于在检测到自动贩售机的柜门关闭前，根据重力传感器感知的货架减少的重量信息和增加的重量信息以及货架在补货完成时初始摆放的单个货品的重量，确定用户是否将欲购买的货品放回原始货架，在未放回原始货架时发出提示，在已放回原始货架后，对该放回的货品不进行结算。
在一些实施例中，还包括：理货模块1050，用于根据自动贩售机内的视频监控图像，检测自动贩售机的货架上的货品摆放是否规则，在不规则时发出提示。
在一些实施例中，还包括：补货模块1060，用于检测到补货员打开自动贩售机的柜门的操作；根据重力传感器感知的货架的重量变化信息以及货架对应摆放的单个货品的重量，确定补充货品的数量；检测到自动贩售机的柜门关闭后，自动记录或上报补充货品的数量。
在一些实施例中，还包括：信息推荐模块1070，用于如果识别到用户的身份信息，根据用户的身份信息对应的购物行为信息为用户进行信息推荐；如果未识别到用户的身份信息，根据人脸识别技术识别用户的性别和年龄中的至少一项，根据识别的用户的性别和年龄中的至少一项为用户进行信息推荐。
图11为本公开自动售货装置一些实施例的结构示意图。
如图11所示，该实施例的装置1100包括：存储器1110以及耦接至该存储器1110的处理器1120，处理器1120被配置为基于存储在存储器1110中的指令，执行前述任意一个实施例中的自动售货方法。
其中，存储器1110例如可以包括系统存储器、固定非易失性存储介质等。系统存储器例如存储有操作系统、应用程序、引导装载程序(Boot Loader)以及其他程序等。
装置1100还可以包括输入输出接口1130、网络接口1140、存储接口1150等。这些接口1130，1140，1150以及存储器1110和处理器1120之间例如可以通过总线1160连接。其中，输入输出接口1130为显示器、鼠标、键盘、触摸屏等输入输出设备提供连接接口。网络接口1140为各种联网设备提供连接接口。存储接口1150为SD卡、U盘等外置存储设备提供连接接口。
本公开还提出一种计算机可读存储介质，其上存储有计算机程序，该程序被处理器执行时实现前述任一个实施例中的自动售货方法。
本领域内的技术人员应当明白，本公开的实施例可提供为方法、系统、或计算机程序产品。因此，本公开可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。而且，本公开可采用在一个或多个其中包含有计算机可用程序代码的计算机可用非瞬时性存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。
本公开是参照根据本公开实施例的方法、设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解为可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器，使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上，使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。
以上所述仅为本公开的较佳实施例，并不用以限制本公开，凡在本公开的精神和原则之内，所作的任何修改、等同替换、改进等，均应包含在本公开的保护范围之内。

Claims (17)

  1. 一种自动售货方法，包括：
    检测到用户打开自动贩售机的柜门的操作；
    判断所述自动贩售机的货架上摆放的是否为预设货品；
    在所述自动贩售机的货架上摆放的是预设货品的情况下，根据货架的重量变化信息确定所述用户欲购买的货品信息；
    在所述自动贩售机的货架上摆放有非预设货品的情况下，根据所述自动贩售机内的视频监控图像确定所述用户欲购买的货品信息；
    检测到所述自动贩售机的柜门关闭后，自动对所述用户欲购买的货品信息进行结算。
  2. 如权利要求1所述的方法，其中，判断所述自动贩售机的货架上摆放的是否为预设货品包括：
    将所述自动贩售机内当前摆放的货品图像与所述自动贩售机内补货完成时初始摆放的货品图像进行比较，如果当前摆放的货架上的货品与初始摆放的该货架上的货品的种类不一致，则判定所述自动贩售机的该货架上摆放有非预设货品。
  3. 如权利要求1所述的方法，其中，判断所述自动贩售机的货架上摆放的是否为预设货品包括：
    将所述自动贩售机内当前摆放的货架上的单个货品的重量与所述自动贩售机内补货完成时初始摆放的该货架上的单个货品的重量进行比较，如果重量不一致，则判定所述自动贩售机的该货架上摆放有非预设货品。
  4. 如权利要求1-3任一项所述的方法，其中，根据货架的重量变化信息确定所述用户欲购买的货品信息包括：
    根据重力传感器感知的重量变化信息以及所述重力传感器对应的货架信息，确定发生重量变化的货架及其重量变化信息；
    根据发生重量变化的货架的重量变化信息，结合发生重量变化的货架承载的货品和单个货品的重量，确定所述用户购买的货品以及货品的数量。
  5. 如权利要求1-3任一项所述的方法，其中，根据所述自动贩售机内的视频监控图像确定所述用户欲购买的货品信息包括：
    从所述自动贩售机内的视频监控图像中定位所述用户欲购买的货品的图像；
    根据预先建立的货品特征库，识别所述用户欲购买的货品的图像对应的货品信息。
  6. 如权利要求5所述的方法，其中，
    定位所述用户欲购买的货品的图像包括：将所述自动贩售机内的视频监控图像输入货品定位模型，所述货品定位模型输出所述用户欲购买的货品的图像，所述货品定位模型通过多个货品的图像及其货品位置的标注信息进行训练得到；
    或者，
    识别所述用户欲购买的货品的图像对应的货品信息包括：将所述用户欲购买的货品的图像输入货品识别模型，所述货品识别模型输出所述用户欲购买的货品的图像对应的货品特征，所述货品识别模型通过多个货品的图像及其货品分类的标注信息进行训练得到，将所述货品识别模型输出的货品特征与所述货品特征库中的货品特征进行匹配，并将匹配到的所述货品特征库中的货品特征对应的货品分类的标注信息作为所述用户欲购买的货品的图像对应的货品信息。
  7. 如权利要求1所述的方法，还包括：
    在检测到所述自动贩售机的柜门关闭前，根据重力传感器感知的货架减少的重量信息和增加的重量信息以及货架在补货完成时初始摆放的单个货品的重量，确定所述用户是否将欲购买的货品放回原始货架，在未放回原始货架时发出提示，在已放回原始货架后，对该放回的货品不进行结算。
  8. 如权利要求1所述的方法，还包括：
    根据所述自动贩售机内的视频监控图像，检测所述自动贩售机的货架上的货品摆放是否规则，在不规则时发出提示。
  9. 如权利要求1所述的方法，还包括：
    检测到补货员打开自动贩售机的柜门的操作；
    根据重力传感器感知的货架的重量变化信息以及货架对应摆放的单个货品的重量，确定补充货品的数量；
    检测到所述自动贩售机的柜门关闭后，自动记录或上报补充货品的数量。
  10. 如权利要求1所述的方法，还包括：
    如果识别到所述用户的身份信息，根据所述用户的身份信息对应的购物行为信息为所述用户进行信息推荐；
    如果未识别到所述用户的身份信息，根据人脸识别技术识别所述用户的性别和年 龄中的至少一项，根据识别的所述用户的性别和年龄中的至少一项为所述用户进行信息推荐。
  11. 一种自动售货装置，包括：
    执行权利要求1-10中任一项所述的自动售货方法的模块。
  12. 一种自动售货装置，包括：
    存储器；以及
    耦接至所述存储器的处理器，所述处理器被配置为基于存储在所述存储器中的指令，执行权利要求1-10中任一项所述的自动售货方法。
  13. 一种计算机可读存储介质，其上存储有计算机程序，该程序被处理器执行时实现权利要求1-10中任一项所述的自动售货方法。
  14. 一种自动贩售机，包括：
    工控机；
    重力传感器，设置于所述自动贩售机的每层货架的下表面；
    内置摄像头，设置于每层货架对应的所述自动贩售机的侧面内壁，拍摄方向对准货架区域；
    电磁锁，设置于所述自动贩售机的柜门处，能够打开或关闭所述柜门；
    其中，所述工控机与所述重力传感器、所述内置摄像头、所述电磁锁分别电连接，所述工控机在用户身份被有效识别的情况下，向所述电磁锁发出指令打开所述柜门，并将所述重力传感器感知的货架的重量变化信息和所述内置摄像头拍摄的视频监控图像传送至后台的自动售货装置，以供所述自动售货装置确定所述用户欲购买的货品信息，并将所述柜门关闭的信息传送至所述自动售货装置，以触发所述自动售货装置自动对所述用户欲购买的货品信息进行结算。
  15. 如权利要求14所述的自动贩售机，其中，
    所述内置摄像头在相邻两层货架对应的所述自动贩售机的两侧面内壁交错设置；
    或者，所述重力传感器包括重力敏感元件；或者，所述重力传感器包括重力敏感元件和数字变换元件，所述重力敏感元件与所述数字变换元件电连接；所述重力敏感元件用于感知货架承载货品的重量信息；所述数字变换元件用于将货品的重量信息变换为货品的数量信息。
  16. 如权利要求14所述的自动贩售机，还包括：
    外置摄像头和触摸屏中的至少一个；
    所述外置摄像头，用于拍摄用户的面部图像，并传送至所述自动售货装置，以供所述自动售货装置识别所述用户的身份；
    所述触摸屏，用于展示所述自动售货装置定义的内容，或者，用于接收并展示后台的所述自动售货装置向所述用户推荐的信息。
  17. 一种自动售货系统，包括：权利要求14-16任一项所述的自动贩售机和权利要求11-12任一项所述的自动售货装置。
PCT/CN2019/075147 2018-03-02 2019-02-15 自动售货方法和系统以及自动售货装置和自动贩售机 WO2019165895A1 (zh)

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
CN201820294180.2 2018-03-02
CN201820294180.2U CN208433046U (zh) 2018-03-02 2018-03-02 自动贩售机和自动售货系统
CN201810175168.4A CN108460908A (zh) 2018-03-02 2018-03-02 自动售货方法和系统以及自动售货装置和自动贩售机
CN201810175168.4 2018-03-02

Publications (1)

Publication Number Publication Date
WO2019165895A1 true WO2019165895A1 (zh) 2019-09-06

Family

ID=67805173

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/075147 WO2019165895A1 (zh) 2018-03-02 2019-02-15 自动售货方法和系统以及自动售货装置和自动贩售机

Country Status (1)

Country Link
WO (1) WO2019165895A1 (zh)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111696264A (zh) * 2020-06-19 2020-09-22 李德政 一种具有自动识别功能的智能终端及运营系统和运营方法
CN112634524A (zh) * 2020-12-18 2021-04-09 江苏生花汇享科技有限公司 一种称重式无人售货设备的称重纠错方法
CN112734446A (zh) * 2020-12-31 2021-04-30 北京每日优鲜电子商务有限公司 基于视觉和重力检测的商品验证方法和装置
CN113128391A (zh) * 2021-04-15 2021-07-16 广东便捷神科技股份有限公司 一种基于人脸识别的无人售货机上货取货的方法
CN113763629A (zh) * 2021-03-25 2021-12-07 北京京东乾石科技有限公司 一种智能售货柜以及检测异物的方法

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO1996038798A1 (en) * 1995-06-01 1996-12-05 Brun, Heidi, M. A vending machine, a vending system and methods for operating same
CN106781014A (zh) * 2017-01-24 2017-05-31 广州市蚁道互联网有限公司 自动售货机及其运行方法
US20170301173A1 (en) * 2014-09-23 2017-10-19 Bubbly Group As Vending machine
CN107301749A (zh) * 2017-07-12 2017-10-27 杨智勇 一种自动支付方法
CN107730782A (zh) * 2017-10-19 2018-02-23 安徽小豆网络科技有限公司 基于视觉图像识别的自助收银机的无人便利店运营系统
CN108460908A (zh) * 2018-03-02 2018-08-28 北京京东尚科信息技术有限公司 自动售货方法和系统以及自动售货装置和自动贩售机
CN208433046U (zh) * 2018-03-02 2019-01-25 北京京东尚科信息技术有限公司 自动贩售机和自动售货系统

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO1996038798A1 (en) * 1995-06-01 1996-12-05 Brun, Heidi, M. A vending machine, a vending system and methods for operating same
US20170301173A1 (en) * 2014-09-23 2017-10-19 Bubbly Group As Vending machine
CN106781014A (zh) * 2017-01-24 2017-05-31 广州市蚁道互联网有限公司 自动售货机及其运行方法
CN107301749A (zh) * 2017-07-12 2017-10-27 杨智勇 一种自动支付方法
CN107730782A (zh) * 2017-10-19 2018-02-23 安徽小豆网络科技有限公司 基于视觉图像识别的自助收银机的无人便利店运营系统
CN108460908A (zh) * 2018-03-02 2018-08-28 北京京东尚科信息技术有限公司 自动售货方法和系统以及自动售货装置和自动贩售机
CN208433046U (zh) * 2018-03-02 2019-01-25 北京京东尚科信息技术有限公司 自动贩售机和自动售货系统

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111696264A (zh) * 2020-06-19 2020-09-22 李德政 一种具有自动识别功能的智能终端及运营系统和运营方法
CN112634524A (zh) * 2020-12-18 2021-04-09 江苏生花汇享科技有限公司 一种称重式无人售货设备的称重纠错方法
CN112634524B (zh) * 2020-12-18 2022-07-08 江苏生花汇享科技有限公司 一种称重式无人售货设备的称重纠错方法
CN112734446A (zh) * 2020-12-31 2021-04-30 北京每日优鲜电子商务有限公司 基于视觉和重力检测的商品验证方法和装置
CN113763629A (zh) * 2021-03-25 2021-12-07 北京京东乾石科技有限公司 一种智能售货柜以及检测异物的方法
CN113128391A (zh) * 2021-04-15 2021-07-16 广东便捷神科技股份有限公司 一种基于人脸识别的无人售货机上货取货的方法
CN113128391B (zh) * 2021-04-15 2024-02-06 广东便捷神科技股份有限公司 一种基于人脸识别的无人售货机上货取货的方法

Similar Documents

Publication Publication Date Title
WO2019165895A1 (zh) 自动售货方法和系统以及自动售货装置和自动贩售机
WO2019165892A1 (zh) 自动售货方法、装置和计算机可读存储介质
CN108875664B (zh) 选购商品的识别方法、装置以及售货机
CN108460908A (zh) 自动售货方法和系统以及自动售货装置和自动贩售机
WO2019154087A1 (zh) 无人售货设备及其售货方法和无人售货系统
US20200258070A1 (en) Purchased product checkout support system
RU2739542C1 (ru) Система автоматической регистрации для торговой точки
JP6342039B1 (ja) 商品を管理するためのシステム、方法、及びプログラム
WO2019161734A1 (zh) 无人售货设备及其售货方法和无人售货系统
US20200065791A1 (en) Self-service product return using computer vision and artificial intelligence
US11526843B2 (en) Product identification systems and methods
WO2018137136A1 (zh) 自动售货机及其运行方法
JP7264401B2 (ja) 会計方法、装置及びシステム
EP4075399A1 (en) Information processing system
JP2018206372A (ja) 商品を管理するためのシステム、方法、及びプログラム
CN111126990B (zh) 物品自动识别方法、结算方法、装置、终端及存储介质
CN109190706A (zh) 无人售货方法、装置及系统
CN107945392B (zh) 自动售货机及采集证据的方法以及存储介质
CN108171286B (zh) 无人售货方法及其系统
CN208433046U (zh) 自动贩售机和自动售货系统
CN109993891A (zh) 称重式智能货柜的控制方法
JP6375926B2 (ja) 商品認識装置及び商品登録装置
Hsu et al. New generation artificial intelligent vending machine system based on LoRaWan IOT network
US11651416B2 (en) Goods purchase analysis assist system
US20190156317A1 (en) Container, shopping assistance system for use with same, shopping assistance method for use with same

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: 19760373

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: 19760373

Country of ref document: EP

Kind code of ref document: A1