WO2022042507A1 - 商品盘点方法和自动售货机 - Google Patents

商品盘点方法和自动售货机 Download PDF

Info

Publication number
WO2022042507A1
WO2022042507A1 PCT/CN2021/114166 CN2021114166W WO2022042507A1 WO 2022042507 A1 WO2022042507 A1 WO 2022042507A1 CN 2021114166 W CN2021114166 W CN 2021114166W WO 2022042507 A1 WO2022042507 A1 WO 2022042507A1
Authority
WO
WIPO (PCT)
Prior art keywords
cargo lane
target
cargo
lane
commodity
Prior art date
Application number
PCT/CN2021/114166
Other languages
English (en)
French (fr)
Inventor
张连峰
朱先刚
邢月启
周萧
姜天信
Original Assignee
威海新北洋数码科技有限公司
山东新北洋信息技术股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 威海新北洋数码科技有限公司, 山东新北洋信息技术股份有限公司 filed Critical 威海新北洋数码科技有限公司
Publication of WO2022042507A1 publication Critical patent/WO2022042507A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G07—CHECKING-DEVICES
    • G07F—COIN-FREED OR LIKE APPARATUS
    • G07F11/00—Coin-freed apparatus for dispensing, or the like, discrete articles
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/40—Scenes; Scene-specific elements in video content
    • G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/50—Context or environment of the image
    • G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
    • G—PHYSICS
    • G07—CHECKING-DEVICES
    • G07F—COIN-FREED OR LIKE APPARATUS
    • G07F11/00—Coin-freed apparatus for dispensing, or the like, discrete articles
    • G07F11/02—Coin-freed apparatus for dispensing, or the like, discrete articles from non-movable magazines
    • G07F11/04—Coin-freed apparatus for dispensing, or the like, discrete articles from non-movable magazines in which magazines the articles are stored one vertically above the other
    • G07F11/16—Delivery means
    • 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
    • G07F9/10—Casings or parts thereof, e.g. with means for heating or cooling

Definitions

  • the present application relates to the technical field of self-service vending, for example, to a commodity inventory method and an automatic vending machine.
  • the related art discloses an automatic vending machine with an automatic inventory function.
  • the automatic vending machine includes a control device and a cabinet.
  • the cabinet is provided with a cargo bucket and a plurality of cargo lanes.
  • the surface of the cabinet is provided with a pickup port.
  • the cargo lane In order to store a plurality of commodities arranged in sequence along the length direction of the cargo lane, the cargo lane has an outlet, a push plate for pushing the commodities to move is arranged in the cargo lane, and the cargo bucket is arranged to be opposite to any one of the cargo lanes, and receives the goods from the cargo lane.
  • the goods sent out are exported and delivered to the pick-up port.
  • a camera and a laser ranging sensor are arranged on the cargo hopper.
  • the camera is set to take an image of the goods in the cargo lane.
  • the control device identifies the type of the goods through the image.
  • the laser ranging sensor is set In order to measure the total length of the commodities in the cargo lane by measuring the distance from the push plate to the laser ranging sensor, the control device calculates the quantity of the commodities in the cargo lane according to the type and total length of the commodities in the cargo lane.
  • the present application provides a commodity inventory method and an automatic vending machine, which can improve the problem of low accuracy of commodity inventory in the related art.
  • a commodity inventory method is provided, which is applied to an automatic vending machine.
  • the automatic vending machine includes a cabinet, and a plurality of cargo lanes for storing commodities are arranged in the cabinet.
  • the mechanism is set to transport the goods in the cargo lane to the exit of the cargo lane, and the commodity inventory method includes:
  • the target commodity is a commodity in the target cargo lane that is closest to the exit of the target cargo lane;
  • An automatic vending machine including a control device and a cabinet, a plurality of cargo lanes for storing commodities are arranged in the cabinet, the cargo lane has an outlet for outputting the commodities, and the cargo lane is provided with a conveying mechanism, and the conveying mechanism is set to supply the goods.
  • the exit of the channel conveys the goods in the cargo channel, and the control device is set to:
  • the target commodity is a commodity in the target cargo lane that is closest to the exit of the target cargo lane;
  • FIG. 1 is a schematic diagram of an automatic vending machine in an embodiment of the application
  • FIG. 2 is an internal schematic diagram of an automatic vending machine in an embodiment of the application
  • FIG. 3 is a block diagram of the composition of an automatic vending machine in an embodiment of the application.
  • FIG. 4 is a schematic diagram of a cargo lane in an embodiment of the application.
  • FIG. 5 is a schematic diagram of a cargo bucket in an embodiment of the application.
  • FIG. 6 is a schematic diagram of the cooperation of a cargo bucket and a cargo channel in an embodiment of the application
  • Fig. 7 is an enlarged view of part VII in Fig. 6;
  • FIG. 8 is a flowchart of a method for inventorying commodities in an embodiment of the present application.
  • FIG. 9 is a schematic diagram of an image of a target cargo lane in an embodiment of the present application.
  • FIG. 10 is a schematic diagram of the distribution of test marks in a preset cargo lane according to an embodiment of the application.
  • FIG. 11 is a schematic diagram of an image of a cargo lane captured by a camera in an embodiment of the present application.
  • Icons 010-automatic vending machine; 100-cabinet; 110-cabinet; 120-cabinet door; 121-human-computer interaction device; ;212-Second partition; 220-Push plate; 221-Detection part; 222-Transmission gear; 223-Pulley; 224-Transmission belt; device; 310-cargo bucket; 311-cargo inlet; 312-outlet; 313-drive gear; 314-drive motor; 315-cargo roller; 320-drive mechanism; 400-camera; 500-ranging sensor; 600 - Controls.
  • the staff when the staff replenishes a cargo lane, they are often used to placing the commodities to be replenished in the cargo lane from the inside to the outside.
  • the frontmost commodity in the cargo lane will be a certain distance from the exit of the cargo lane.
  • the distance between the frontmost commodity in the cargo aisle (that is, the commodity closest to the exit of the cargo aisle) and the camera may be too large, resulting in the recognition of the image captured by the camera.
  • the type of commodity is missed or misidentified, or when the total length of the commodity is measured by measuring the distance from the push plate to the laser distance sensor through the laser distance sensor, the total length of the measured commodity is inaccurate. , which will affect the accuracy of product inventory.
  • the embodiments of the present application provide a commodity inventory method and an automatic vending machine.
  • the commodities in the target cargo channel are moved to the outlet of the cargo channel, so that the frontmost commodity is located in the target cargo channel. Exit, and then take inventory.
  • the commodities in the target cargo lane are counted, the commodities can be closer to the camera, the ranging sensor, etc., so as to improve the accuracy of image recognition and the accuracy of ranging, thereby improving the accuracy of the inventory.
  • an automatic vending machine provided by an embodiment of the present application is first introduced below.
  • the vending machine 010 provided by the embodiment of the present application includes a cabinet 100 , a cargo lane 200 disposed in the cabinet 100 , a pickup device 300 , a camera 400 , a ranging sensor 500 , and a control device 600 .
  • the cargo aisle 200 is configured to store commodities, and can realize the delivery of commodities from the cargo aisle 200; the pickup device 300 is configured to deliver the commodities delivered from the cargo aisle 200 to the user; the camera 400 is configured to collect images in the cabinet 100, The image of the cargo lane 200 and the images of the commodities in the cargo lane 200 are included; the ranging sensor 500 is configured to detect the length of the commodities arranged in the cargo lane 200 to calculate the quantity of the commodities in the cargo lane 200 .
  • the cargo lane 200 , the pickup device 300 , the camera 400 , and the ranging sensor 500 are all electrically connected to the control device 600 .
  • the control device 600 can control a plurality of devices to perform corresponding operations.
  • the cabinet 100 includes a cabinet body 110 and a cabinet door 120 , the cabinet body 110 has a horizontally forward opening, and the cabinet door 120 is rotatably connected to the cabinet body 110 to open or close the opening of the cabinet body 110 .
  • a pickup port 122 is provided on the cabinet door 120 , and the pickup port 122 communicates with the inside and the outside of the cabinet 100 , so that a user can take out the goods in the pickup device 300 in the cabinet 100 through the pickup port 122 .
  • a pickup door can also be set at the pickup port 122 to open or close the pickup port 122, so that when the vending machine 010 is idle, the pickup port 122 can be kept closed to prevent snakes, insects, rats, ants, etc.
  • a human-computer interaction device 121 is also provided on the cabinet door 120, and the human-computer interaction device 121 is configured to receive instructions input by the user (such as purchase, query, parameter configuration and other related instructions) and send information to the user (such as commodities on sale). information, advertisement information, prompt information, etc.), the human-computer interaction device 121 is electrically connected to the control device 600 .
  • the human-computer interaction device 121 includes a touch screen, which can satisfy the input and output of information at the same time.
  • the human-computer interaction device 121 may further include a keyboard, a mouse, a speaker, and the like.
  • the vending machine 010 includes a plurality of cargo lanes 200 , and the multiple cargo lanes 200 are arranged in the cabinet 110 in a distributed manner of multiple rows and multiple columns, and each cargo lane 200 is located in the front and rear of the vending machine 010 It extends in the direction (that is, the direction indicated by the arrow ef in the figure).
  • One end of the cargo aisle 200 close to the cabinet door 120 has an exit, and commodities can be delivered to the cargo aisle 200 from the exit of the cargo aisle 200 to reach the pickup device 300 .
  • a certain space is reserved between the exit of the cargo aisle 200 and the cabinet door 120, and the pickup device 300 is arranged in the space.
  • one cargo channel 200 is configured to store one kind of commodity, and the commodities in the cargo channel 200 are arranged along the length direction of the cargo channel 200 .
  • FIG. 4 is a schematic diagram of a cargo channel 200 according to an embodiment of the present application.
  • each cargo channel 200 includes a bottom plate 210 , a first partition plate 211 and a second partition plate 212 .
  • the bottom plate 210 is configured to support commodities, and the first partition plate 211 and the second partition plate 212 are oppositely disposed on the bottom plate 210 along the width direction of the cargo channel 200 (ie, the direction indicated by the arrow ab in the figure), and the two are parallel to each other.
  • the bottom plate 210, An accommodation space of the cargo channel 200 is formed between the first partition plate 211 and the second partition plate 212 .
  • Multiple cargo lanes 200 in the same row (layer) can share a bottom plate 210; the accommodation spaces of two adjacent cargo lanes 200 can be separated by a partition between the two cargo lanes 200.
  • the first partition 211 can simultaneously serve as the second partition 212 of another adjacent cargo lane 200 .
  • the left baffle of a cargo lane 200 is defined as the first baffle 211 of the cargo lane 200
  • the right baffle is defined as the second baffle 212 of the cargo lane 200 .
  • each cargo lane 200 is provided with a conveying mechanism, and the conveying mechanism is configured to send the commodities in the cargo lane 200 from the outlet of the cargo lane 200 .
  • the conveying mechanism includes a transmission component and a push plate 220 disposed in the cargo channel 200.
  • the transmission component is configured to drive the push plate 220 to move in the cargo channel 200.
  • the push plate 220 moves, it pushes the commodities in the cargo channel 200 along the The cargo lane 200 moves in the longitudinal direction.
  • the position of the push plate 220 in the cargo channel 200 is after the last commodity, so it can push all commodities in the cargo channel 200 to move toward the exit of the cargo channel 200 together, and the commodities located at the front end (that is, , a commodity closest to the exit of the cargo lane 200 ) will first reach the exit of the cargo lane 200 and be sent to the cargo lane 200 .
  • the transmission assembly includes a transmission gear 222, a pulley 223, a transmission belt 224 and a driven pulley (not shown in the figure).
  • the transmission gear 222 and the pulley 223 are coaxially and fixedly connected, the pulley 223 and the driven pulley are arranged at intervals in the length direction of the cargo channel 200, the transmission belt 224 is sleeved between the pulley 223 and the driven pulley, and the transmission belt 224 is connected to the push plate 220. connect.
  • the transmission gear 222 rotates
  • the pulley 223 rotates with the transmission gear 222 at the same angular velocity, thereby driving the transmission belt 224 to move.
  • the merchandise moves toward the exit of the cargo lane 200 .
  • the transmission assembly is disposed on the first partition 211 of the cargo lane 200 corresponding to the transmission assembly, and most of the transmission belt 224 and the driven wheels are embedded in the first partition 211, which makes the structure compact and reduces space occupation .
  • the push plate 220 is provided with a detection part 221 .
  • the detection part 221 is the detection object of the distance measuring sensor 500 , and the distance measuring sensor 500 can detect the total length of the commodities in the cargo lane 200 by measuring the distance between the detection part 221 and the distance measuring sensor 500 .
  • the detection part 221 is located above the first partition 211 of the cargo lane 200 corresponding to the detection part 221 .
  • FIG. 5 is a schematic diagram of the cargo bucket 310 in an embodiment of the application
  • FIG. 6 is a schematic diagram of the cooperation between the cargo bucket 310 and the cargo channel 200 in an embodiment of the application
  • FIG. 7 is an enlarged view of part VII in FIG. 6 .
  • the pickup device 300 is disposed in the space between the exit of the cargo channel 200 and the cabinet door 120 , and the pickup device 300 includes a cargo bucket 310 and is configured to drive the cargo bucket 310
  • the drive mechanism 320 moves between the pickup port 122 and the exit of any one of the cargo lanes 200 .
  • the driving mechanism 320 can drive the bucket 310 to the exit of any one of the cargo lanes 200 to receive the commodities sent from the cargo lane 200 and transport the commodities to the position corresponding to the pickup port 122 for the user to take away.
  • the driving mechanism 320 may include a horizontal driving mechanism and a vertical driving mechanism to move the cargo bucket 310 to the exit of any one of the cargo lanes 200.
  • the horizontal drive mechanism is configured to drive the vertical drive mechanism and the bucket 310 to move in the horizontal direction together
  • the vertical drive mechanism is configured to drive the bucket 310 to move in the vertical direction.
  • the cargo hopper 310 has a cargo inlet 311 and an outlet 312 , wherein the cargo inlet 311 is opposite to the outlet of the cargo channel 200 , and the outlet 312 is opposite to the cabinet door 120 .
  • Commodities sent from the cargo channel 200 enter the cargo hopper 310 from the cargo inlet 311 of the cargo hopper 310 .
  • the cargo hopper 310 moves to the position of the delivery port 122 , the user can pass the cargo hopper 122 and the delivery port 312 to the cargo hopper. Items in 310 are taken out.
  • a feeding roller 315 is provided at the cargo inlet 311 of the cargo hopper 310, and the feeding roller 315 can be driven by the motor (in the figure) provided on the cargo hopper 310. (not shown) is driven so as to transport the commodities located at the loading port 311 into the interior of the cargo bucket 310 .
  • the side of the cargo bucket 310 facing the cargo lane 200 is further provided with a drive gear 313 and a drive motor 314 configured to drive the drive gear 313 to rotate.
  • the drive gear 313 on the cargo bucket 310 can engage with the drive gear 222 of the transmission assembly of the cargo lane 200, so the drive motor 314 can pass the drive gear 313, the drive gear 222 .
  • the transmission belt 224 finally drives the push plate 220 of the cargo channel 200 to move, so as to realize the delivery of commodities from the outlet of the cargo channel 200 . Commodities sent out from the outlet of the cargo lane 200 can fall into the cargo hopper 310 .
  • the push plates 220 in the cargo lane 200 adopt this driving method, so that the push plates 220 of all cargo lanes 200 are driven by the same drive motor 314 and the same drive gear 313 , thereby reducing the cost by reducing the use of the drive motor 314 and the space occupied in the cabinet 100 .
  • a drive motor 314 may also be provided in each cargo lane 200 to drive the push plate 220 to move.
  • the camera 400 is disposed in the cargo compartment 310 .
  • the ranging sensor 500 is also disposed in the cargo bucket 310 . Both the camera 400 and the ranging sensor 500 are located on the side of the cargo bucket 310 facing the cargo lane 200 . By arranging the camera 400 and the ranging sensor 500 on the cargo bucket 310 , when the cargo bucket 310 moves to the exit position of the target cargo channel 200 , the camera 400 can acquire the image of the target cargo channel 200 in a targeted manner, and measure the distance.
  • the sensor 500 can be opposed to the detection part 221 on the push plate 220 of the target cargo lane 200 , so as to obtain the distance between the push plate 220 of the target cargo lane 200 and the ranging sensor 500 .
  • the ranging sensor 500 may be a laser sensor, and by emitting laser light and receiving the laser light reflected by the detecting portion 221 , the position between the detecting portion 221 and the ranging sensor 500 is detected.
  • the image captured by the camera 400 should include the image of the target cargo lane 200, and may also include the image of the area around the target cargo lane 200; when the cargo bin 310 moves to the target cargo lane At the exit of 200 , the distance measuring sensor 500 should be opposite to the detection part 221 of the push plate 220 of the target cargo channel 200 in the length direction of the target cargo channel 200 .
  • the ranging sensor 500 may also be fixedly arranged at the exit of the cargo lane 200 , and each cargo lane 200 is provided with a ranging sensor 500 ; The exit is even provided on the inner side of the cabinet door 120 .
  • control device 600 is set to:
  • the target commodity is a commodity in the target cargo lane 200 that is closest to the exit of the target cargo lane 200;
  • the commodity moves a first distance so that the target commodity is located at the exit of the target cargo lane 200 ; the commodities in the target cargo lane 200 are counted.
  • the control device 600 controls the bucket 310 to move to the exit of the target cargo aisle 200, and controls the camera 400 to capture an image of the target cargo aisle 200, and the image should contain the image of the target commodity.
  • the control device 600 calculates the first distance between the exit of the target cargo channel 200 (ie the front end of the bottom plate 210 ) and the target commodity by analyzing and processing the image of the target cargo channel 200 collected by the camera 400 .
  • the first distance can be considered as a distance in which no commodities are displayed in the target cargo lane 200 .
  • the target commodities in the target cargo lane 200 when the commodities in the target cargo lane 200 cannot occupy the entire cargo lane 200, the target commodities are far away from the camera 400, so it is difficult to accurately identify the types of commodities in the cargo lane 200 through the camera 400; Moreover, since the total length of the commodities in the target cargo lane 200 is obtained by subtracting the distance between the ranging sensor 500 and the target commodity from the distance between the ranging sensor 500 and the push plate 220 , the commodities in the target cargo lane 200 If the entire cargo lane 200 cannot be filled, the position of the target commodity is uncertain, and the distance between the ranging sensor 500 and the target commodity is also uncertain, which makes it difficult to determine the total number of commodities in the target cargo lane 200. Length is accurately detected.
  • the push plate 220 is controlled to move the first distance, so that the commodities in the target cargo lane 200 move the first distance, so that the target commodities are located at the exit of the target cargo lane 200 .
  • the distance between the target product and the camera 400 is shortened, so that the camera 400 can capture the image of the product with higher definition, and then the control device 600 can more accurately obtain the feature information of the product through the image captured by the camera 400, so as to accurately Determine the type of the commodity, and at the same time, when the target commodity is located at the exit of the target cargo lane 200, since the target commodity is located at a certain position, the measurement result of the ranging sensor 500 can be accurately calculated to obtain the commodity in the target cargo lane 200.
  • the total length is equal to the distance between the distance measuring sensor 500 and the detection part 221 minus the fixed distance from the distance measuring sensor 500 to the exit of the target cargo aisle 200 . Therefore, it is possible to accurately count the commodities in the target cargo aisle 200 .
  • the image of each type of commodity and the length of a single commodity can be stored in the memory of the vending machine 010 in advance. Therefore, the type of commodity can be obtained through the collected commodity image, and then the length of the single commodity corresponding to the commodity can be obtained. .
  • the distance measuring sensor 500 can detect the distance between the distance measuring sensor 500 and the detection part 221 of the push plate 220 . Since the push plate 220 is in contact with the rear end of the last commodity in the cargo lane 200 , the distance measuring sensor 500 and the detection part 221 The distance between can be regarded as the distance between the last commodity in the cargo lane 200 and the distance measuring sensor 500 .
  • the distance between the distance measuring sensor 500 and the exit of the cargo channel 200 in the length direction of the cargo channel 200 is a fixed value
  • the distance between the exit of the target cargo channel 200 and the push plate 220 can pass the actual measurement value of the distance measuring sensor 500 Calculated by subtracting this fixed value. Since the target commodity has moved to the exit of the target cargo aisle 200 before the distance between the push plate 220 and the exit of the target cargo aisle 200 is calculated, the distance between the exit of the target cargo aisle 200 and the push plate 220 can be used as The total length of the commodities in the target cargo lane 200. Then, the total length of the commodities in the target cargo lane 200 is divided by the length of a single commodity to obtain the quantity of commodities in the target cargo lane 200 .
  • the distance measuring sensor 500 used in this embodiment is a laser distance measuring sensor, and a push plate 220 with a detection portion 221 is arranged in the conveying mechanism. By measuring the distance between the distance measuring sensor 500 and the push plate 220, the cargo lane is calculated. The total length of the item within 200. It is an optional embodiment of the present application to adopt such a distance measuring sensor and set the push plate 220 with the detection part 221, and adopt corresponding detection and calculation methods. In other embodiments of the present application, the distance measuring sensor The type of 500, the setting method of the conveying mechanism, and the method for detecting and calculating the total length of the commodity can be replaced. For example, the following embodiment is adopted: the ranging sensor 500 is an image sensor, and a ruler is provided on the cargo lane 200.
  • the image of the last commodity and the ruler at the position of the commodity is used to determine the position of the commodity in the target cargo lane 200 through image analysis. Since the position of the target commodity at the front end has been determined (located at the exit of the target freight aisle 200), the total length of the commodity in the target freight aisle 200 can also be accurately calculated by the position of the end commodity, and then the goods can be calculated. The number of goods in the road.
  • FIG. 8 is a flowchart of a commodity inventory method in an embodiment of the present application.
  • an embodiment of the present application provides a method for inventorying commodities, which can be applied to the vending machine 010 provided by the embodiment of the present application.
  • the commodity inventory method includes the following steps.
  • Step S100 Obtain a first distance from the target commodity in the target cargo lane to the exit of the target cargo lane, where the target commodity is a commodity in the target cargo lane that is closest to the exit of the target cargo lane.
  • obtaining the first distance from the target commodity in the target cargo lane 200 to the exit of the target cargo lane 200 may include: controlling the camera 400 to photograph the cargo lane The image of 200 ; the first distance from the target commodity in the target cargo lane 200 to the exit of the target cargo lane 200 is calculated according to the image of the cargo lane 200 .
  • the control device 600 first controls the drive mechanism 320 to drive the bucket 310 to the exit of the target cargo lane 200 according to the received instruction (for example, the inventory instruction input by the user through the human-computer interaction device 121 ), and then controls the camera 400 to photograph the cargo lane 200 .
  • the image of the cargo lane 200 should contain the image of the target cargo lane 200 . Then, the image of the cargo channel 200 is analyzed to obtain the distance between the exit of the target cargo channel 200 (that is, the front end of the bottom plate 210 ) and the target commodity.
  • FIG. 9 is a schematic diagram of an image of a target cargo lane 200 in an embodiment of the present application.
  • calculating the first distance from the target commodity in the target cargo channel 200 to the exit of the target cargo channel 200 according to the image of the cargo channel 200 may include: determining the distance between the target commodity and the target cargo channel 200 according to the image of the cargo channel 200 A plurality of pixel points P on the intersection of the bottom plate 210; according to each pixel point P, the second distance L between a target commodity and the exit of the target cargo lane 200 is obtained; The minimum value of the two distances L is used as the first distance from the target commodity to the exit of the target cargo lane 200 .
  • the intersection line between the target commodity and the bottom plate 210 of the target cargo lane 200 is the line formed by the intersection of the target commodity and the bottom plate 210 of the target cargo lane 200 .
  • the intersection line may be a straight line or a curved line.
  • the intersection line is a straight line
  • the intersection line is a curved line.
  • the intersection line between the target commodity and the bottom plate 210 of the target cargo lane 200 is composed of pixel points P, and each pixel point P on the intersection line in the image corresponds to a pixel point on the actual target commodity. physical point.
  • the distance from the front end of the physical point corresponding to the pixel point P along the length direction of the cargo lane 200 to the bottom plate 210 is calculated, and this distance is taken as the corresponding distance of the pixel point P.
  • the second distance L between the target commodity and the exit of the target freight lane 200 is calculated, and this distance is taken as the corresponding distance of the pixel point P.
  • the image of the cargo lane 200 captured by the camera 400 is a distorted image (as shown in FIG. 11 ). Therefore, the distance between the pixel point P in the image and the front end of the bottom plate 210 cannot truly represent the pixel point P The distance between the corresponding physical point on the actual target commodity and the actual front end of the bottom plate 210 . For example, for the pixel point P closest to the front end of the bottom plate 210 on the image, the physical point corresponding to the pixel point P on the actual target commodity is not necessarily the closest physical point to the front end of the bottom plate 210 .
  • the distance from the front end of the physical point corresponding to the pixel point P along the length direction of the cargo lane 200 to the front end of the bottom plate 210 is obtained, and this distance is used as the The second distance L between the target commodity corresponding to the pixel point P and the exit of the target cargo lane 200, since the second distance L can truly reflect the distance from the physical point on the actual target commodity to the front end of the bottom plate 210 of the cargo lane 200, Therefore, by determining the minimum value of the plurality of second distances L as the first distance from the target commodity to the exit of the target cargo lane 200 , the first distance can accurately reflect the edge where the target commodity intersects with the bottom plate 210 of the target cargo lane 200 .
  • the distance from the foremost front end to the exit of the target cargo lane 200 By driving the commodities in the target cargo lane 200 to move the first distance, the front end of the target commodities can be located at the exit of the target cargo lane 200, thereby not only ensuring that the target commodities do not fall from the target cargo lane 200, but also improving the The accuracy of identification and the accuracy of the measurement of the total length of the commodity, thereby improving the accuracy of commodity inventory.
  • the second distance L between a target commodity and the exit of the target cargo lane 200 is calculated according to each pixel point P, which may include:
  • this embodiment uses different conversion polynomials. to calculate the second distance L corresponding to the pixel points P at different positions, to obtain the second distance L that can more accurately reflect the real situation. Due to image distortion, the transformation polynomials used by pixels P at different positions may be different.
  • the conversion polynomial includes an input parameter and an output value
  • the input parameter includes the coordinate position of the pixel point P along the width direction of the cargo lane 200
  • the output value is the second distance L corresponding to the pixel point P.
  • the position of the pixel point P in the image of the cargo lane 200 along the width direction of the cargo lane 200 means that the pixel point P is located on the bottom plate 210 between the first partition plate 211 and the second partition plate 212 The upper position along the width direction of the cargo lane 200 .
  • FIG. 9 is only a schematic diagram of the image of the cargo lane 200.
  • the straight line outline of the actual cargo lane 200 is also represented as a straight line in the figure; and in the image actually collected by the camera 400, the straight line in the figure may be Distorted and bent.
  • FIG. 10 is a schematic diagram of the distribution of test marks in the preset cargo lane 200 according to an embodiment of the present application.
  • the conversion comparison table may be obtained in advance in the following manner.
  • the camera 400 is controlled to capture an image of the preset cargo lane 200, wherein the preset cargo lane 200 is provided with a test mark, the test mark is located on the bottom plate 210 of the preset cargo lane 200, and the test mark includes M first dividing lines 230 and N second dividing lines 240 , the first dividing lines 230 extend along the length direction of the preset cargo lane 200 , the second dividing lines 240 extend along the width direction of the preset cargo lane 200 , M first dividing lines 230 and N
  • the second dividing line 240 divides the bottom plate 210 of the preset cargo lane 200 into a plurality of areas; for each first dividing line 230, a plurality of intersections on the first dividing line 230 are selected, Assuming the position in the image of the cargo lane 200 and the actual position of the intersection, the conversion polynomial of the first dividing line 230 is obtained, where the intersection refers to the point where the first dividing line 230 and the second dividing line 240 intersect;
  • the preset cargo lane 200 may be the target cargo lane 200 , or may be another cargo lane 200 having the same width as the target cargo lane 200 .
  • the conversion polynomial of the pixel point P which may include:
  • the first dividing line 230 closest to the pixel point P is determined; the closest first dividing line is determined according to the pre-stored conversion comparison table 230 corresponds to the conversion polynomial, and the conversion polynomial corresponding to the nearest first dividing line 230 is used as the conversion polynomial of the pixel point P.
  • the position of the pixel point P along the width direction of the cargo lane 200 in the image of the cargo lane 200 and the position of each first dividing line 230 along the width direction of the cargo lane 200 in the image of the preset cargo lane 200 determine the distance The first dividing line 230 closest to the pixel point P. Since the image of the cargo lane 200 captured by the camera 400 is a distorted image, by using test marks for pre-calibration, it can be obtained that for different positions along the width direction of the cargo lane 200, the effect of image distortion on the distance conversion, that is, , to obtain the conversion polynomials at different positions.
  • the conversion polynomial can be used to calculate the pixel point P according to the coordinate position of the pixel point P in the image to obtain the physical point corresponding to the pixel point P on the actual target product.
  • the second distance L from the point to the foremost end of the bottom plate 210 of the target cargo lane 200 ensures the accuracy of the calculation of the second distance L.
  • the expected value of the distance between the selected first dividing line 230 closest to the pixel point P and the pixel point P is smaller, then adopt The conversion polynomial of the first dividing line 230 calculates the pixel point P, and the obtained second distance L is more accurate.
  • the test mark can be set on the bottom plate 210 of the preset cargo lane 200 in the form of sticking, or can be set on the bottom plate 210 of the preset cargo lane 200 in the form of printing.
  • the width of the cargo channel 200 can be adjusted, that is, the distance between the first partition 211 and the second partition 212 can be adjusted, before the vending machine 010 leaves the factory, or in the cargo channel of the vending machine 010
  • the preset cargo lane 200 needs to be calibrated with test marks, and a conversion comparison table corresponding to the width of the preset cargo lane 200 is obtained.
  • the test marks are set on the bottom plate 210 of the cargo channel 200 in the form of stickers. After the width of the cargo channel 200 is adjusted, the staff can make a 200 is suitable for the width of the test mark, the test mark is pasted on the bottom plate 210 of the cargo channel 200, and the conversion comparison table of the cargo channel 200 is obtained by passing the test mark.
  • determining a plurality of pixel points P on the intersection of the target commodity and the bottom plate 210 of the target cargo lane 200 according to the image of the cargo lane 200 may further include:
  • the image of the cargo lane 200 is analyzed and processed to determine the detection area of the target cargo lane 200 in the image of the cargo lane 200 , wherein the detection area of the target cargo lane 200 refers to the area defined by the edge of the bottom plate 210 of the target cargo lane 200 ; Carry out line-by-line detection of the pixel point P of the detection area along the set direction, determine the mutation point in the detection area, and use the mutation point in the detection area as the pixel on the intersection of the target commodity and the bottom plate 210 of the target cargo lane 200 point P.
  • the detection area of the target cargo lane 200 determined from the image of the cargo lane 200 is often not a regular rectangle.
  • other areas in the image of the cargo lane 200 other than the detection area of the target cargo lane 200 may be shielded, or, image segmentation may be used to extract shipments
  • the detection area of the target cargo lane 200 in the lane 200 image may be shielded, or, image segmentation may be used to extract shipments.
  • the interference of other abrupt change points other than the detection area of the target freightway 200 in the image of the freightway 200 can be excluded, and the detected abrupt change point can be guaranteed to be a pixel point on the intersection of the target commodity and the bottom plate 210 of the target freightway 200 P.
  • one or more complete cargo channels 200 may be included in the image of the cargo channel 200 .
  • the image captured by the camera 400 includes only one complete cargo lane 200, that is, the target cargo lane 200; when the width of the target cargo lane 200 is narrow, the camera 400 In the captured images, in addition to the target cargo lane 200 , other complete cargo lanes 200 may also be included.
  • the camera 400 is disposed on the cargo bucket 310.
  • FIG. 11 is a schematic diagram of an image of a cargo lane captured by a camera in an embodiment of the present application. As shown in FIG.
  • the camera 400 takes pictures based on the first partition 211 located on the left side of the bottom plate 210 of the target cargo channel 200 , that is, in the acquired image of the cargo channel 200 , the first partition The left apex angle of the front end of 211 is located at the edge of the image of the cargo lane 200.
  • the image of the cargo lane 200 is analyzed and processed to determine the detection area of the target cargo lane 200 in the image of the cargo lane 200, which may include:
  • the coordinate positions of multiple positioning points in the image of the cargo lane 200 are pre-defined, and after obtaining the width of the target cargo lane 200 , the cargo lane 200 with the width is obtained by looking up a table. The coordinate positions of the corresponding multiple positioning points are then drawn according to the multiple positioning points to obtain the detection area of the target cargo lane 200 .
  • the image of the target cargo lane 200 captured by the camera 400 is captured based on a partition of the target cargo lane 200 . For example, as shown in FIG. 11 , the left side of the front end of the first partition 211 The side top angle is located at the edge of the image of the cargo channel 200. In this way, after the width of the cargo channel 200 is determined, the position of the cargo channel 200 in the image of the cargo channel 200 is also determined.
  • the multiple positioning points corresponding to the width of 1 can quickly and accurately obtain the detection area of the target cargo lane 200 .
  • the image of the cargo lane 200 is analyzed and processed to determine the detection area of the target cargo lane 200 in the image of the cargo lane 200, which may include:
  • the image of the cargo lane 200 is segmented by using a deep learning algorithm, and the partition area of the target cargo lane 200 in the image of the cargo lane 200 is extracted;
  • the adjacent distance fitting lines corresponding to the front ends of each partition are used as reference lines to calculate the rear ends of the partitions.
  • the detection area of the target cargo lane 200 is obtained by calculating the front end and the front end and the rear end of the second partition 212 .
  • the adjacent distance fitting line corresponding to the front end of each partition refers to the first dividing line 230 in the target cargo aisle 200 that is closest to the front end of the partition. As shown in FIG.
  • a test mark may be set in each target cargo lane 200 , and the test mark includes a plurality of first dividing lines 230 extending along the length direction of the target cargo lane 200 .
  • the leftmost first dividing line 230 in the target cargo lane 200 can be used as the adjacent distance fitting line of the first partition 211, and the adjacent distance fitting line can be used as the reference line
  • the rear end of the first partition 211 is calculated according to the front end of the first partition 211
  • the rightmost first dividing line 230 in the target cargo lane 200 is used as the adjacent distance fitting line of the second partition 212 .
  • the distance fitting line is taken as the reference line
  • the rear end of the second separator 212 is calculated according to the front end of the second separator 212 .
  • an image segmentation algorithm based on deep learning is used, sample images of various commodities are marked in advance, and feature files are generated by training according to the sample images of various commodities.
  • a deep learning algorithm is used to perform image segmentation processing, and the commodity area and the partition area of the target cargo lane 200 are extracted.
  • step S200 the commodities in the target cargo lane are driven to move a first distance, so that the target commodities are located at the exit of the target cargo lane.
  • the control device 600 controls the drive motor 314 on the cargo bucket 310 to rotate, and drives the drive motor 314 to rotate.
  • the gear 313 , the transmission gear 222 , the pulley 223 and the transmission belt 224 move to drive the push plate 220 to move the first distance to the exit of the target cargo lane 200 , and finally make the target commodity move to the exit of the target cargo lane 200 .
  • the commodities in the target cargo lane 200 have been as close to the exit of the target cargo lane 200 as possible, that is, the commodities are as close to the camera 400 as possible, when the camera 400 recognizes the type of commodities in the target cargo lane 200, the higher accuracy.
  • the target commodity is located at the exit of the cargo lane 200, and the position of the target commodity is deterministic. Therefore, the distance between the ranging sensor 500 and the detection unit 221 obtained by the ranging sensor 500 is used to calculate the commodity in the target cargo lane 200. more accurate results can also be obtained.
  • step S300 the commodities in the target cargo lane are counted.
  • inventorying the commodities in the target cargo lane includes detecting the type and quantity of commodities in the target cargo lane 200 .
  • the control device 600 can calculate the distance between the target cargo lane 200 and the the quantity of the item.
  • the image of each type of item and the length of a single item may be pre-stored in the memory of the vending machine 010.
  • the control device 600 can compare it with the image in the commodity image library in the memory to determine the type of the commodity, and then acquire the length of a single commodity corresponding to the commodity.
  • the distance between the distance measuring sensor 500 and the detection part 221 can be regarded as the distance between the last commodity in the cargo lane 200 and the distance measuring sensor 500 . Since the distance between the distance measuring sensor 500 and the outlet of the cargo channel 200 in the length direction of the cargo channel 200 is a fixed value, the distance between the exit of the target cargo channel 200 and the push plate 220 can pass the measured value of the distance measuring sensor 500 Calculated by subtracting this fixed value.
  • the target commodity Before calculating the distance between the push plate 220 and the exit of the target cargo lane 200, the target commodity has been moved to the exit of the target cargo lane 200, so the distance between the exit of the target cargo lane 200 and the push plate 220 can be As the total length of the commodities in the target cargo lane 200 . It can be seen that moving the target commodity to the exit of the target cargo lane 200 in step S200 is the premise of accurately calculating the total length of the commodities in the cargo lane 200 . After calculating the total length of the commodities in the target cargo lane 200, divide the total length of the commodities in the target cargo lane 200 by the length of a single item to obtain the quantity of the commodities in the target cargo lane 200. Inventory of commodities within Dao 200.
  • the commodity inventory method can obtain the first distance from the target commodity in the target cargo channel to the exit of the target cargo channel before the inventory of the commodities in the target cargo channel, and drive the commodities in the target cargo channel to move the first distance.
  • the target commodity is located at the exit of the target cargo lane by a distance, which shortens the distance between the target commodity and the camera, which is beneficial to improve the accuracy of commodity type identification through the image of the cargo lane;
  • a definite position is also more conducive to improving the accuracy of measuring the total length of the goods in the cargo lane through the ranging sensor, thereby improving the accuracy of the automatic inventory of the vending machine.
  • the self-service vending machine provided in the embodiment of the present application is used to implement the above-mentioned method for inventorying commodities, so the accuracy of automatic inventorying is high.

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Computational Linguistics (AREA)
  • Software Systems (AREA)
  • Control Of Vending Devices And Auxiliary Devices For Vending Devices (AREA)

Abstract

提供了一种商品盘点方法和自动售货机。商品盘点方法包括：获取目标货道中目标商品到目标货道的出口的第一距离，其中，目标商品为目标货道中离目标货道的出口最近的一个商品（S100），并驱动目标货道中的商品移动第一距离使目标商品位于目标货道的出口处（S200），对目标货道中的商品进行盘点（S300）。

Description

商品盘点方法和自动售货机
本申请要求在2020年08月31日提交中国专利局、申请号为202010899381.7的中国专利申请的优先权，该申请的全部内容通过引用结合在本申请中。
技术领域
本申请涉及自助售卖技术领域，例如涉及一种商品盘点方法和自动售货机。
背景技术
随着自动售货技术的发展，很多自动售货机均具有自动盘点功能。相关技术公开了一种具有自动盘点功能的自动售货机，该自动售货机包括控制装置和柜体，柜体内设置有货斗和多个货道，柜体表面设置有取货口，货道设置为存储沿货道的长度方向依次排布的多个商品，货道具有出口，货道中设置有用于推动商品移动的推板，货斗设置为能够与任意一个货道相对，接收由货道的出口送出的商品并将商品运送至取货口，货斗上设置有摄像头和激光测距传感器，摄像头设置为拍摄货道内商品的图像，控制装置通过该图像识别商品的类型，激光测距传感器设置为通过测量推板到激光测距传感器的距离来测量货道内商品的总长度，控制装置根据货道内商品的类型以及总长度计算货道内商品的数量。
对于相关技术的自动售货机，存在商品识别不准确或者测量的商品的总长度不准确，导致商品盘点的准确性较低的问题。
发明内容
本申请提供一种商品盘点方法和自动售货机，能够改善相关技术中商品盘点的准确性较低的问题。
提供一种商品盘点方法，应用于自动售货机，自动售货机包括机柜，机柜内设置有用于存储商品的多个货道，货道具有用于输出商品的出口，且货道设置有输送机构，输送机构设置为向货道的出口输送货道内的商品，商品盘点方法包括：
获取目标货道中的目标商品到目标货道的出口的第一距离，其中，目标商品为目标货道中离目标货道的出口最近的一个商品；
驱动目标货道中的商品移动第一距离，使目标商品位于目标货道的出口处；
对目标货道中的商品进行盘点。
还提供一种自动售货机，包括控制装置和机柜，机柜内设置有用于存储商品的多个货道，货道具有用于输出商品的出口，且货道设置有输送机构，输送机构设置为向货道的出口输送货道内的商品，控制装置设置为：
获取目标货道中的目标商品到目标货道的出口的第一距离，其中，目标商品为目标货道中离目标货道的出口最近的一个商品；
驱动目标货道中的商品移动第一距离，使目标商品位于目标货道的出口处；
对目标货道中的商品进行盘点。
附图说明
图1为本申请一种实施例中自动售货机的示意图；
图2为本申请一种实施例中自动售货机的内部示意图；
图3为本申请一种实施例中自动售货机的组成框图；
图4为本申请一种实施例中货道的示意图；
图5为本申请一种实施例中货斗的示意图；
图6为本申请一种实施例中货斗与货道配合的示意图；
图7为图6中局部VII的放大图；
图8为本申请一种实施例中商品盘点方法的流程图；
图9为本申请一种实施例中目标货道的图像的示意图；
图10为本申请一种实施例中预设货道中测试标记的分布示意图；
图11为本申请一种实施例中摄像头拍摄的货道的图像的示意图。
图标：010-自动售货机；100-机柜；110-柜体；120-柜门；121-人机交互装置；122-取货口；200-货道；210-底板；211-第一隔板；212-第二隔板；220-推板；221-检测部；222-传动齿轮；223-带轮；224-传动带；230-第一分割线；240-第二分割线；300-取货装置；310-货斗；311-入货口；312-出货口；313-驱动齿轮；314-驱动电机；315-入货辊；320-驱动机构；400-摄像头；500-测距传感器；600-控制装置。
具体实施方式
下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行描述，所描述的实施例是本申请一部分实施例，而不是全部的实施例。通常在此处附图中描述和示出的本申请实施例的组件可以以多种不同的配置来布置和设计。
因此，以下对在附图中提供的本申请的实施例的描述并非旨在限制要求保护的本申请的范围，而是仅仅表示本申请的选定实施例。
相似的标号和字母在下面的附图中表示类似项，因此，一旦一项在一个附图中被定义，则在随后的附图中不需要对该项进行定义和解释。
在本申请的描述中，若出现术语“上”、“下”、“内”、“外”等指示的方位或位置关系为基于附图所示的方位或位置关系，或者是该产品使用时惯常摆放的方位或位置关系，仅是为了便于描述本申请和简化描述，而不是指示或暗示所指的装置或元件必须具有特定的方位、以特定的方位构造和操作，因此不能理解为对本申请的限制。
此外，若出现术语“第一”、“第二”等仅用于区分描述，而不能理解为指示或暗示相对重要性。
在不冲突的情况下，本申请的实施例中的特征可以相互结合。
相关技术的自动售货机，在工作人员对一个货道进行补货时，往往习惯在货道中由内向外依次摆放要补入的商品，工作人员补货后经常会出现所补的商品不能占满整个货道的情况，在这种情况下，货道中最前端的商品距离货道的出口就会有一定距离。当自动售货机在对货道中的商品进行盘点时，可能由于货道中最前端的商品(也即，离货道出口最近的商品)与摄像头之间的距离过大，导致通过摄像头拍摄的图像识别商品的类型时出现漏识或误识的情况，或者，通过激光测距传感器测量推板到激光测距传感器的距离来测量商品的总长度时，出现所测量的商品的总长度不准确的情况，这就会影响商品盘点的准确性。
因此，本申请实施例提供一种商品盘点方法和自动售货机，在对目标货道进行商品盘点前，通过将目标货道内的商品向货道出口移动，使得最前端的商品位于目标货道的出口处，然后再进行盘点。通过这种方式，使得在对目标货道的商品进行盘点时，商品能够更靠近摄像头、测距传感器等，以提高图像识别的准确性以及测距的准确性，从而提高盘点的准确率。为了方便理解本申请提供的商品盘点方法，下面首先对本申请一种实施例提供的自动售货机进行介 绍。
图1为本申请一种实施例中自动售货机010的示意图；图2为本申请一种实施例中自动售货机010的内部示意图；图3为本申请一种实施例中自动售货机010的组成框图。请参考图1至图3，本申请实施例提供的自动售货机010包括机柜100以及设置于机柜100内的货道200、取货装置300、摄像头400、测距传感器500以及控制装置600。货道200设置为存储商品，并能够实现将商品从货道200内送出；取货装置300设置为将从货道200内送出的商品输送给用户；摄像头400设置为采集机柜100内的图像，包括货道200的图像以及货道200内商品的图像；测距传感器500设置为检测货道200内所排列的商品的长度，以计算货道200内商品的数量。货道200、取货装置300、摄像头400、测距传感器500均与控制装置600电连接。控制装置600可以控制多个装置执行相应的作业。
本实施例中，机柜100包括柜体110和柜门120，柜体110具有水平朝前的开口，柜门120可转动地连接于柜体110，以设置为打开或者关闭柜体110的开口。柜门120上设置有取货口122，取货口122连通机柜100的内部和外部，使得用户可以通过取货口122取走机柜100内的取货装置300中的商品。可选的，取货口122处还可以设置取货门，来实现取货口122的打开或者关闭，使得自动售货机010在闲置时，取货口122能够保持关闭以防止蛇虫鼠蚁等动物或者其他异物进入自动售货机010的机柜100内。可选的，柜门120上还设置有人机交互装置121，人机交互装置121设置为接收用户输入的指令(比如购买、查询、参数配置等相关指令)以及向用户发送信息(比如在售商品信息、广告信息、提示信息等)，人机交互装置121与控制装置600电连接。在本实施例中，人机交互装置121包括触控屏，可同时满足信息的输入和输出。在可选的其他实施例中，人机交互装置121还可以包括键盘、鼠标、扬声器等。
在本实施例中，自动售货机010包括多个货道200，多个货道200以多行多列的分布方式设置在柜体110内，每一个货道200均在自动售货机010的前后方向(即图中箭头ef所指方向)上延伸。货道200靠近柜门120的一端具有出口，商品可以从货道200的出口送出货道200，而到达取货装置300。货道200的出口与柜门120之间预留一定的空间，取货装置300设置于该空间内。在本实施例中，一个货道200设置为存储一种商品，货道200内的商品沿着货道200的长度方向排列。
图4为本申请一种实施例中货道200的示意图，如图4所示，每个货道200包括底板210、第一隔板211和第二隔板212。底板210设置为支撑商品，第一隔板211和第二隔板212沿货道200的宽度方向(即图中箭头ab所指方向)相 对设置于底板210上且二者相互平行，底板210、第一隔板211和第二隔板212之间形成货道200的容纳空间。同一行(层)的多个货道200可以共用一个底板210；相邻两个货道200的容纳空间可以通过两个货道200之间的一个隔板隔开，因此，一个货道200的第一隔板211可以同时作为相邻的另一个货道200的第二隔板212。在本实施例中，将一个货道200的左侧的隔板定义为该货道200的第一隔板211，将右侧的隔板定义为该货道200的第二隔板212。
请继续参见图4，在本实施例中，每一个货道200均设置有输送机构，输送机构设置为将货道200内的商品从货道200的出口送出。在本实施例中，输送机构包括传动组件和设置于货道200中的推板220，传动组件设置为带动推板220在货道200中移动，推板220移动时推动货道200中的商品沿货道200的长度方向移动。在本实施例中，推板220在货道200中的位置，位于最后一个商品之后，因此能够推动该货道200内所有的商品一同朝向货道200的出口移动，位于前端的商品(也即，离货道200的出口最近的一个商品)会首先到达货道200的出口而被送出货道200。
在本实施例中，传动组件包括传动齿轮222、带轮223、传动带224以及从动轮(图中未示出)。传动齿轮222与带轮223同轴地固定连接，带轮223和从动轮在货道200的长度方向上间隔设置，传动带224套设在带轮223和从动轮之间，传动带224与推板220连接。当传动齿轮222转动时，带轮223随传动齿轮222以相同的角速度转动，从而带动传动带224运动，传动带224带动推板220在货道200内沿着货道200的长度方向移动，以实现推动商品向货道200的出口移动。在本实施例中，传动组件设置于传动组件对应的货道200的第一隔板211，并且传动带224的大部分、从动轮均嵌入在第一隔板211内部，使得结构紧凑，减少空间占用。
在本实施例中，推板220上设置有检测部221。检测部221作为测距传感器500的检测对象，测距传感器500可以通过测量检测部221与测距传感器500之间的距离，从而检测货道200中商品的总长度。在本实施例中，检测部221位于检测部221对应的货道200的第一隔板211上方。
图5为本申请一种实施例中货斗310的示意图；图6为本申请一种实施例中货斗310与货道200配合的示意图；图7为图6中局部VII的放大图。请参照图5至图7，在本实施例中，取货装置300设置在货道200的出口与柜门120之间的空间内，取货装置300包括货斗310和设置为驱动货斗310在取货口122与任意一个货道200的出口之间移动的驱动机构320。驱动机构320能够将货斗310驱动至任意一个货道200的出口处，以接收该货道200送出的商品，并将该商品输送到取货口122对应的位置，供用户取走。可选的，驱动机构320可以 包括水平驱动机构和竖直驱动机构，来实现将货斗310移动到任意一个货道200的出口处。比如，水平驱动机构设置为驱动竖直驱动机构和货斗310共同在水平方向上移动，竖直驱动机构设置为驱动货斗310在竖直方向上移动。
在本实施例中，货斗310具有入货口311和出货口312，其中，入货口311与货道200的出口相对，出货口312与柜门120相对。由货道200送出的商品从货斗310的入货口311进入货斗310，当货斗310移动到取货口122的位置处，用户可通过取货口122、出货口312将货斗310内的商品取出。为了使得从货道200送出的商品能够顺利进入到货斗310内，货斗310的入货口311处设置有入货辊315，入货辊315可以被货斗310上设置的电机(图中未示出)驱动，从而将位于入货口311处的商品运送到货斗310内部。
在本实施例中，货斗310朝向货道200的一侧还设置有驱动齿轮313以及设置为驱动驱动齿轮313转动的驱动电机314。当货斗310移动到一个货道200的出口时，货斗310上的驱动齿轮313能够与该货道200的传动组件的传动齿轮222啮合，因此，驱动电机314能够通过驱动齿轮313、传动齿轮222、传动带224，最终带动该货道200的推板220移动，从而实现将商品从货道200的出口送出。从货道200的出口送出的商品能够落入货斗310内。货道200内的推板220采用这种驱动方式，使得所有货道200的推板220通过同一个驱动电机314和同一个驱动齿轮313来驱动，通过减少驱动电机314的使用，从而降低了成本以及机柜100内的空间占用。在可选的其他实施例中，也可以在每个货道200内均设置一个驱动电机314，来驱动推板220移动。
为了使摄像头400能够更好地获取货道200中商品的图像，以对商品的类型进行判断，在本申请实施例中，摄像头400设置于货斗310。同理，测距传感器500也设置于货斗310。摄像头400和测距传感器500均位于货斗310朝向货道200的一侧。通过将摄像头400和测距传感器500设置在货斗310上，使得货斗310在移动到目标货道200的出口位置处时，摄像头400能够有针对性地获取目标货道200的图像，测距传感器500能够与目标货道200的推板220上的检测部221相对，从而获取目标货道200的推板220与测距传感器500之间的距离。测距传感器500可以是激光传感器，通过发射激光，并接收检测部221反射的激光，来检测检测部221与测距传感器500之间的位置。当货斗310移动到目标货道200的出口处时，摄像头400采集的图像应当包含目标货道200的图像，也可以包含目标货道200周围区域的图像；当货斗310移动到目标货道200的出口处时，测距传感器500应当与目标货道200的推板220的检测部221在目标货道200的长度方向上相对。在本申请可选的其他实施例中，测距传感器500也可以固定设置在货道200的出口处，每一个货道200分别设置一个 测距传感器500；摄像头400也可以设置在货道200的出口处，甚至设置在柜门120的内侧。
在本申请实施例中，控制装置600设置为：
获取目标货道200中的目标商品到目标货道200的出口的第一距离；其中，目标商品为目标货道200中离目标货道200的出口最近的一个商品；驱动目标货道200中的商品移动第一距离，使目标商品位于目标货道200的出口处；对目标货道200中的商品进行盘点。
控制装置600控制货斗310移动到目标货道200的出口处，控制摄像头400采集目标货道200的图像，该图像中应该包含目标商品的图像。控制装置600通过对摄像头400所采集的目标货道200的图像进行分析处理后计算出目标货道200的出口(即底板210的前端)与目标商品之间的第一距离。该第一距离可以认为是目标货道200内没有陈列商品的一段距离。相关技术中，由于在目标货道200中的商品不能占满整个货道200的情况下，目标商品距离摄像头400较远，因此难以通过摄像头400对货道200内商品的类型进行准确地识别；并且，由于目标货道200内商品的总长度通过测距传感器500与推板220之间的距离减去测距传感器500与目标商品之间的距离得到，因此，在目标货道200中的商品不能占满整个货道200的情况下，目标商品的位置具有不确定性，进而测距传感器500与目标商品之间的距离也不确定，这就会导致了难以对目标货道200内商品总长度进行准确地检测。但本申请中，在获取了第一距离之后，控制推板220移动第一距离，令目标货道200中的商品移动第一距离，使得目标商品位于目标货道200的出口处。这样缩短了目标商品与摄像头400的距离，因此能够使摄像头400拍摄到清晰度更高的商品的图像，进而使控制装置600通过摄像头400拍摄的图像能够更准确地获取商品的特征信息，从而准确确定商品的类型，同时，当目标商品位于目标货道200的出口处时，由于目标商品位于一个确定的位置，因此，根据测距传感器500的测量结果能够准确计算得到目标货道200内商品的总长度，该长度等于测距传感器500与检测部221之间的距离减去测距传感器500到目标货道200的出口的固定距离，因此，能够实现对目标货道200内商品进行准确地盘点。
每一个类型的商品的图像以及单件商品的长度可以预先存储在自动售货机010的存储器内，因此，可以通过采集的商品图像来获取商品的类型，进而获取该商品对应的单件商品的长度。测距传感器500能够检测测距传感器500与推板220的检测部221之间的距离，由于推板220抵接于该货道200最后一个商品的后端，因此测距传感器500与检测部221之间的距离可以看作是该货道200中最后一个商品与测距传感器500之间的距离。由于测距传感器500与货道200 出口在货道200的长度方向上的间距是一个固定值，因此，目标货道200的出口与推板220之间的距离可以通过测距传感器500的实测值减去该固定值计算出来。由于在计算推板220与目标货道200的出口之间的距离之前，目标商品已经移动到目标货道200的出口处，因此目标货道200的出口与推板220之间的距离即可以作为目标货道200内商品的总长度。然后再用目标货道200内商品的总长度除以单件商品的长度，即可得到目标货道200内商品的数量。
本实施例所采用的测距传感器500为激光测距传感器，并在输送机构中设置具有检测部221的推板220，通过测量测距传感器500与推板220之间的距离，进而计算货道200内商品的总长度。采用这种测距传感器并设置具有检测部221的推板220，并采用相应的检测、计算方法，是本申请的一种可选的实施方式，在本申请的其他实施方式中，测距传感器500的类型、输送机构的设置方式以及检测、计算商品总长度的方法可以进行替换，比如采用以下实施方式：测距传感器500为图像传感器，货道200上设置有标尺，通过图像传感器采集货道中最靠后的一个商品与该商品所在位置的标尺的图像，通过图像分析确定该商品在目标货道200中的位置。由于处于最前端的目标商品的位置已经确定(位于目标货道200的出口处)，因此，通过最后端的商品的位置，也可以精确地计算目标货道200内商品的总长度，进而计算该货道内商品的数量。
图8为本申请一种实施例中商品盘点方法的流程图。请参照图8并结合图1至图7，本申请实施例提供一种商品盘点方法，可以应用于本申请实施例提供的自动售货机010。商品盘点方法包括以下步骤。
步骤S100，获取目标货道中的目标商品到目标货道的出口的第一距离，其中，目标商品为目标货道中离目标货道的出口最近的一个商品。
以本申请实施例提供的自动售货机010为例，在本实施例中，获取目标货道200中的目标商品到目标货道200的出口的第一距离，可以包括：控制摄像头400拍摄货道200的图像；根据货道200的图像计算目标货道200中的目标商品到目标货道200的出口的第一距离。控制装置600根据接收到的指令(比如用户通过人机交互装置121输入的盘点指令)，先控制驱动机构320将货斗310驱动到目标货道200的出口处，再控制摄像头400拍摄货道200的图像，货道200的图像中应包含有目标货道200的图像。然后对货道200的图像进行分析，获取目标货道200的出口(也即底板210的最前端)与目标商品之间的距离。
图9为本申请一种实施例中目标货道200的图像的示意图。请参照图9，根据货道200的图像计算目标货道200中的目标商品到目标货道200的出口的第 一距离，可以包括：根据货道200的图像确定目标商品与目标货道200的底板210的交线上的多个像素点P；根据每个像素点P计算得到一个目标商品与目标货道200的出口之间的第二距离L；将多个像素点P对应的多个第二距离L中的最小值作为目标商品到目标货道200的出口的第一距离。
目标商品与目标货道200的底板210的交线为目标商品与目标货道200的底板210相交所形成的线，该交线可以为直线，也可以为曲线，比如当目标商品包括长方体的包装盒时，该交线为直线，当目标商品为瓶装饮料时，该交线为曲线。在货道200的图像中，目标商品与目标货道200的底板210的交线是由一个个像素点P组成的，图像中该交线上的每个像素点P对应实际目标商品上的一个物理点。在本实施例中，根据每个像素点P计算得到沿货道200的长度方向，该像素点P对应的物理点到底板210的最前端的距离，并将该距离作为该像素点P对应的目标商品与目标货道200的出口之间的第二距离L。
在一些情况下，摄像头400拍摄到的货道200的图像是畸变图像(如图11所示)，因此，图像中像素点P与底板210的最前端之间的距离不能真实表示该像素点P在实际目标商品上所对应的物理点与实际的底板210的最前端之间的距离。例如，在图像上离底板210的最前端最近的像素点P，该像素点P在实际目标商品上所对应的物理点不一定是离底板210的最前端最近的物理点。因此，本实施例中，可选的，根据每个像素点P计算得到沿货道200的长度方向，该像素点P对应的物理点到底板210的最前端的距离，并将该距离作为该像素点P对应的目标商品与目标货道200的出口之间的第二距离L，由于第二距离L能够真实反映实际目标商品上的物理点到货道200的底板210的最前端的距离，因此，通过将多个第二距离L中的最小值确定为目标商品到目标货道200的出口的第一距离，能够使第一距离准确反映目标商品与目标货道200的底板210相交的边缘的最前端到目标货道200的出口的距离。通过驱动目标货道200中的商品移动第一距离，能够使目标商品的最前端位于目标货道200的出口处，从而既能保证目标商品不从目标货道200中掉落，又能提高商品识别的准确性和商品总长度测量的准确性，进而提高商品盘点的准确性。
根据每个像素点P计算得到一个目标商品与目标货道200的出口之间的第二距离L，可以包括：
确定像素点P在货道200的图像中沿货道200的宽度方向的位置；根据像素点P在货道200的图像中沿货道200的宽度方向的位置以及预先存储的转换对照表，确定像素点P的转换多项式，其中，转换多项式用于根据像素点P的位置计算像素点P对应的第二距离L；根据像素点P在货道200的图像中沿货道200的宽度方向的位置以及像素点P对应的转换多项式，计算像素点P对应 的第二距离L。
由于摄像头400拍摄的图像容易产生畸变，图像中不同位置反映出来的距离或者图像中的尺寸比例，可能与实际情况不符。因此，为了保证针对不同位置的每一个像素点P计算出的第二距离L能够更准确地反映该像素点P对应的物理点到底板210的最前端的距离，本实施例通过不同的转换多项式来计算不同位置的像素点P所对应的第二距离L，来获取能够更准确反映真实情况的第二距离L。由于图像畸变的原因，不同位置的像素点P所采用的转换多项式可能是不同的。
转换多项式包括输入参数和输出值，输入参数包括像素点P沿货道200的宽度方向的坐标位置，输出值为该像素点P对应的第二距离L。通过转换多项式，在根据货道200的图像已知一个像素点P沿货道200的宽度方向的坐标位置的情况下，能够计算得到该像素点P在目标商品上对应的物理点到目标货道200的底板210的最前端的第二距离L。在本实施例中，像素点P在货道200的图像中沿货道200的宽度方向的位置是指，该像素点P在位于第一隔板211和第二隔板212之间的底板210上沿货道200的宽度方向的位置。
图9仅为货道200的图像的一种示意图，为了方便展示，将实际货道200的直线轮廓在图中也对应表现为直线；而摄像头400实际采集的图像中，图中的直线条可能发生畸变而弯曲。
图10为本申请一种实施例中预设货道200中测试标记的分布示意图。请参照图10，可选的，转换对照表可以预先通过以下方式获取。
控制摄像头400拍摄预设货道200的图像，其中，预设货道200中设置有测试标记，测试标记位于预设货道200的底板210上，且测试标记包括M条第一分割线230和N条第二分割线240，第一分割线230沿预设货道200的长度方向延伸，第二分割线240沿预设货道200的宽度方向延伸，M条第一分割线230和N条第二分割线240将预设货道200的底板210划分为多个区域；对每条第一分割线230，选取该第一分割线230上的多个交叉点，根据每个交叉点在预设货道200的图像中的位置以及该交叉点的实际位置，得到该第一分割线230的转换多项式，其中，交叉点是指第一分割线230和第二分割线240相交的点；将每条第一分割线230与该第一分割线230的转换多项式的对应关系存储为转换对照表。
通过在每条第一分割线230上选取多个交叉点，根据每个交叉点在预设货道200的图像中的坐标位置以及每个交叉点的实际位置，通过多项式拟合生成该第一分割线230的多项式系数，从而得到该第一分割线230的转换多项式。 其中，预设货道200可以为目标货道200，也可以为与目标货道200的宽度相同的其他货道200。
根据像素点P在货道200的图像中沿货道200的宽度方向的位置以及预先存储的转换对照表，确定像素点P的转换多项式，可以包括：
根据像素点P在货道200的图像中沿货道200的宽度方向的位置，确定离该像素点P最近的第一分割线230；根据预先存储的转换对照表确定该最近的第一分割线230对应的转换多项式，将该最近的第一分割线230对应的转换多项式作为该像素点P的转换多项式。
根据该像素点P在货道200的图像中沿货道200的宽度方向的位置和每条第一分割线230在预设货道200的图像中沿货道200的宽度方向的位置，确定离该像素点P最近的第一分割线230。由于摄像头400拍摄的货道200的图像是畸变图像，因此，通过使用测试标记进行预先标定，能够得到对于沿货道200的宽度方向的不同位置，由于图像畸变对于距离转换产生的影响，也即，得到不同位置的转换多项式，在确定像素点P的转换多项式后，能够根据该像素点P在图像中的坐标位置，利用转换多项式进行计算，得到该像素点P在实际目标商品上对应的物理点到目标货道200的底板210的最前端的第二距离L，保证第二距离L计算的准确性。当测试标记中的第一分割线230越密集，对应的转换多项式越多，则选取的离像素点P最近的第一分割线230与该像素点P之间距离的期望值就越小，那么采用该第一分割线230的转换多项式对该像素点P进行计算，得到的第二距离L更为准确。
测试标记可以以粘贴形式设置于预设货道200的底板210上，也可以以印刷形式设置于预设货道200的底板210上。通常情况下，货道200的宽度可调整，也即，第一隔板211和第二隔板212之间的距离可以调整，在自动售货机010出厂前，或者在自动售货机010的货道200的宽度调整后，需要使用测试标记对预设货道200进行标定，得到与预设货道200的宽度对应的转换对照表。因此，为了适应货道200的宽度可调整这一特征，可选的，测试标记以粘贴形式设置于货道200的底板210上，在货道200的宽度调整后，工作人员可以制作与货道200的宽度相适应的测试标记，将该测试标记粘贴于货道200的底板210上，通过该测试标记，从而获得该货道200的转换对照表。
可选的，根据货道200的图像确定目标商品与目标货道200的底板210的交线上的多个像素点P，还可以包括：
对货道200的图像进行分析处理，确定货道200的图像中目标货道200的检测区域，其中，目标货道200的检测区域是指由目标货道200的底板210的 边缘所限定的区域；沿设定方向对该检测区域的像素点P进行逐行检测，确定检测区域中的突变点，将检测区域中的突变点作为目标商品与目标货道200的底板210的交线上的像素点P。
由于摄像头400拍摄到的货道200图像往往是畸变图像，因此，从货道200的图像中所确定的目标货道200的检测区域往往不是规则的矩形。为了便于处理，在确定货道200图像中目标货道200的检测区域之后，可以屏蔽掉货道200的图像中除目标货道200的检测区域以外的其他区域，或者，采用图像分割提取出货道200图像中目标货道200的检测区域。如此，可以排除货道200的图像中目标货道200的检测区域以外的其他突变点的干扰，能够确保检测所得的突变点为目标商品与目标货道200的底板210的交线上的像素点P。
受摄像头400的视野范围、安装位置以及货道200宽度的影响，在货道200的图像中，可能包括一个或多个完整的货道200。例如，当目标货道200的宽度较宽时，摄像头400所拍摄的图像中仅包括一个完整的货道200，也即，目标货道200；当目标货道200的宽度较窄时，摄像头400所拍摄的图像中，除了目标货道200外，可能还包括其他的完整货道200。本申请实施例中，摄像头400设置于货斗310上，当货斗310位于目标货道200的出口处时，摄像头400的视野范围、安装位置以及货斗310与目标货道200的相对位置关系能够保证摄像头400所拍摄的目标货道200的图像是以目标货道200的一个隔板为基准进行拍摄的。图11为本申请一种实施例中摄像头拍摄的货道的图像的示意图。如图11所示，摄像头400是以目标货道200的位于底板210左侧的第一隔板211为基准进行拍摄的，也即，在获取到的货道200的图像中，第一隔板211的前端的左侧顶角位于货道200的图像的边缘处，通过如此设置，可以保证摄像头400能够拍摄到目标货道200的完整图像，并在对货道200的图像进行分析处理过程中，能够以图像的一侧边缘为基准进行边缘检测，确定目标货道200的检测区域。
在可选的一种实施例中，对货道200的图像进行分析处理，确定货道200的图像中目标货道200的检测区域，可以包括：
获取目标货道200的宽度；根据目标货道200的宽度确定货道200的图像中的多个定位点；根据多个定位点确定货道200的图像中目标货道200的检测区域。
在本实施例中，根据货道200的宽度，预先定义货道200的图像中多个定位点的坐标位置，在获取目标货道200的宽度后，通过查表获得与该宽度的货道200对应的多个定位点的坐标位置，然后根据多个定位点进行绘制，得到目 标货道200的检测区域。如上所述，可选的，摄像头400所拍摄的目标货道200的图像是以目标货道200的一个隔板为基准进行拍摄的，例如图11所示，第一隔板211的前端的左侧顶角位于货道200的图像的边缘处，这样一来，当货道200的宽度确定后，货道200在货道200的图像中的位置也是确定的，因此，根据与目标货道200的宽度对应的多个定位点，能够快速准确地得到目标货道200的检测区域。
在可选的另一种实施例中，对货道200的图像进行分析处理，确定货道200的图像中目标货道200的检测区域，可以包括：
通过预先训练的特征文件，采用深度学习算法对货道200的图像进行分割处理，提取货道200的图像中目标货道200的隔板区域；从隔板区域中确定目标货道200的第一隔板211和第二隔板212的前端，以每个隔板的前端对应的邻近的距离拟合线作为基准线，计算出该隔板的后端，根据第一隔板211的前端及后端、和第二隔板212的前端及后端进行计算得到目标货道200的检测区域。可选的，与每个隔板的前端对应的邻近的距离拟合线是指目标货道200中与该隔板的前端最近的第一分割线230。如图10所示，可在每个目标货道200中设置测试标记，测试标记中包括多条沿目标货道200的长度方向延伸的第一分割线230，在确定第一隔板211和第二隔板212的前端后，可将目标货道200中最左侧的第一分割线230作为第一隔板211的邻近的距离拟合线，以该邻近的距离拟合线作为基准线，根据第一隔板211的前端计算第一隔板211的后端，将目标货道200中最右侧的第一分割线230作为第二隔板212的邻近的距离拟合线，以该邻近的距离拟合线作为基准线，根据第二隔板212的前端计算第二隔板212的后端。
在本实施例中，采用基于深度学习的图像分割算法，预先标注多种商品的样本图像，根据多种商品的样本图像训练生成特征文件。对于待处理的货道200的图像，通过预先训练的特征文件，采用深度学习算法进行图像分割处理，提取到目标货道200的商品区域和隔板区域。
步骤S200，驱动目标货道中的商品移动第一距离，使目标商品位于目标货道的出口处。
以本申请实施例提供的自动售货机010为例，控制装置600在获得了目标商品与目标货道200的出口之间的第一距离之后，控制货斗310上的驱动电机314转动，带动驱动齿轮313、传动齿轮222、带轮223和传动带224运动，从而驱动推板220向目标货道200的出口移动第一距离，最终使得目标商品移动到目标货道200的出口处。由于目标货道200内的商品已经尽可能地靠近目标 货道200的出口，也即，商品尽可能地靠近摄像头400，因此，摄像头400对目标货道200内的商品的类型进行识别时，识别准确性更高。此时，目标商品位于货道200的出口处，目标商品的位置具有确定性，因此，通过测距传感器500所获得的测距传感器500与检测部221之间的距离计算目标货道200内商品的总长度时，也能够得到更为准确的结果。
步骤S300，对目标货道中的商品进行盘点。
可选的，对目标货道中的商品进行盘点包括检测目标货道200中商品的类型和数量。以本申请实施例提供的自动售货机010为例，控制装置600在获取了目标货道200内商品的类型以及测距传感器500与检测部221之间的距离后，可以计算目标货道200内商品的数量。每一个类型的商品的图像以及单件商品的长度可以预先存储在自动售货机010的存储器内。因此，控制装置600在获取了商品的图像后，可以与存储器内的商品图像库中的图像来进行比对，从而确定该商品的类型，进而获取该商品对应的单件商品的长度。测距传感器500与检测部221之间的距离可以看作是该货道200中最后一个商品与测距传感器500之间的距离。由于测距传感器500与货道200出口在货道200的长度方向上的间距是一个固定值，因此，目标货道200的出口与推板220之间的距离可以通过测距传感器500的实测值减去该固定值计算出来。由于在计算推板220与目标货道200的出口之间的距离之前，目标商品已经被移动到目标货道200的出口处，因此目标货道200的出口与推板220之间的距离即可以作为目标货道200内商品的总长度。由此可见，步骤S200中将目标商品移动到目标货道200的出口处，是精确计算该货道200内商品的总长度的前提。在计算得到目标货道200内商品的总长度之后，再用目标货道200内商品的总长度除以单件商品的长度，即可得到目标货道200内商品的数量，实现对该目标货道200内商品的盘点。
综上，本申请实施例提供的商品盘点方法，能够在对目标货道中的商品进行盘点之前获取目标货道中目标商品到目标货道的出口的第一距离，并驱动目标货道中的商品移动第一距离使目标商品位于目标货道的出口处，缩短了目标商品与摄像头的距离，有利于提高通过货道的图像进行商品类型的识别的准确性；并且，盘点商品前，通过使目标商品位于一个确定的位置，也更有利于提高通过测距传感器测量货道中商品的总长度的准确性，进而提高了自动售货机自动盘点的准确性。本申请实施例提供的自助售货机，用于实现上述的商品盘点方法，因此自动盘点准确性高。

Claims (10)

  1. 一种商品盘点方法，应用于自动售货机，所述自动售货机包括机柜，所述机柜内设置有用于存储商品的多个货道，所述货道具有用于输出所述商品的出口，且所述货道设置有输送机构，所述输送机构设置为向所述货道的出口输送所述货道内的商品，所述商品盘点方法包括：
    获取目标货道中的目标商品到所述目标货道的出口的第一距离，其中，所述目标商品为所述目标货道中离所述目标货道的出口最近的一个商品；
    驱动所述目标货道中的商品移动所述第一距离，使所述目标商品位于所述目标货道的出口处；
    对所述目标货道中的商品进行盘点。
  2. 根据权利要求1所述的方法，其中，所述自动售货机的机柜内还设置有摄像头，所述摄像头设置为获取所述货道的图像；所述获取目标货道中的目标商品到所述目标货道的出口的第一距离，包括：
    控制所述摄像头拍摄所述货道的图像；
    根据所述货道的图像计算所述目标货道中的目标商品到所述目标货道的出口的第一距离。
  3. 根据权利要求2所述的方法，其中，所述货道包括用于支撑所述商品的底板，所述根据所述货道的图像计算所述目标货道中的目标商品到所述目标货道的出口的第一距离，包括：
    根据所述货道的图像确定所述目标商品与所述目标货道的底板的交线上的多个像素点；
    根据每个像素点计算得到一个所述目标商品与所述目标货道的出口之间的第二距离；
    将所述多个像素点对应的多个第二距离中的最小值作为所述目标商品到所述目标货道的出口的所述第一距离。
  4. 根据权利要求3所述的方法，其中，所述根据每个像素点计算得到一个所述目标商品与所述目标货道的出口之间的第二距离，包括：
    确定所述像素点在所述货道的图像中沿所述货道的宽度方向的位置；
    根据所述像素点在所述货道的图像中沿所述货道的宽度方向的位置以及预先存储的转换对照表，确定所述像素点的转换多项式，其中，所述转换多项式用于根据所述像素点的位置计算所述像素点对应的第二距离；
    根据所述像素点在所述货道的图像中沿所述货道的宽度方向的位置以及所述像素点对应的转换多项式，计算所述像素点对应的第二距离。
  5. 根据权利要求4所述的方法，其中，所述转换对照表预先通过以下方式获取：
    控制所述摄像头拍摄预设货道的图像，其中，所述预设货道中设置有测试标记，所述测试标记位于所述预设货道的底板上，且所述测试标记包括M条第一分割线和N条第二分割线，所述第一分割线沿所述预设货道的长度方向延伸，所述第二分割线沿所述预设货道的宽度方向延伸，所述M条第一分割线和所述N条第二分割线将所述预设货道的底板划分为多个区域；
    对每条第一分割线，选取所述第一分割线上的多个交叉点，根据每个交叉点在所述预设货道的图像中的位置以及所述交叉点的实际位置，得到所述第一分割线的转换多项式，其中，所述交叉点是指所述第一分割线和所述第二分割线相交的点；
    将所述每条第一分割线与所述每条第一分割线的转换多项式的对应关系存储为所述转换对照表。
  6. 根据权利要求5所述的方法，其中，所述根据所述像素点在所述货道的图像中沿所述货道的宽度方向的位置以及预先存储的转换对照表，确定所述像素点的转换多项式，包括：
    根据所述像素点在所述货道的图像中沿所述货道的宽度方向的位置，确定离所述像素点最近的第一分割线；
    根据所述预先存储的转换对照表确定所述最近的第一分割线对应的转换多项式，将所述最近的第一分割线对应的转换多项式作为所述像素点的转换多项式。
  7. 根据权利要求3所述的方法，其中，所述根据所述货道的图像确定所述目标商品与所述目标货道的底板的交线上的多个像素点，包括：
    对所述货道的图像进行分析处理，确定所述货道的图像中所述目标货道的检测区域，其中，所述目标货道的检测区域是指由所述目标货道的底板的边缘所限定的区域；
    沿设定方向对所述检测区域的像素点进行逐行检测，确定所述检测区域中的突变点，将所述检测区域中的突变点作为所述目标商品与所述目标货道的底板的交线上的像素点。
  8. 根据权利要求7所述的方法，其中，所述对所述货道的图像进行分析处理，确定所述货道的图像中所述目标货道的检测区域，包括：
    获取所述目标货道的宽度；
    根据所述目标货道的宽度确定所述货道的图像中的多个定位点；
    根据所述多个定位点确定所述货道的图像中所述目标货道的检测区域。
  9. 根据权利要求7所述的方法，其中，所述货道还包括第一隔板和第二隔板，所述第一隔板和所述第二隔板沿所述货道的宽度方向相对设置于所述底板上且所述第一隔板和所述第二隔板相互平行，所述底板、所述第一隔板和所述第二隔板之间形成所述货道的容纳空间，所述对所述货道的图像进行分析处理，确定所述货道的图像中所述目标货道的检测区域，包括：
    通过预先训练的特征文件，采用深度学习算法对所述货道的图像进行分割处理，提取所述货道的图像中所述目标货道的隔板区域；
    从所述隔板区域中确定所述目标货道的第一隔板和第二隔板的前端，以每个隔板的前端对应的邻近的距离拟合线作为基准线，计算出所述每个隔板的后端，根据所述第一隔板的前端及后端、和所述第二隔板的前端及后端进行计算得到所述目标货道的检测区域。
  10. 一种自动售货机，包括控制装置和机柜，所述机柜内设置有用于存储商品的多个货道，所述货道具有用于输出所述商品的出口，且所述货道设置有输送机构，所述输送机构设置为向所述货道的出口输送所述货道内的商品，所述控制装置设置为：
    获取目标货道中的目标商品到所述目标货道的出口的第一距离，其中，所述目标商品为所述目标货道中离所述目标货道的出口最近的一个商品；
    驱动所述目标货道中的商品移动所述第一距离，使所述目标商品位于所述目标货道的出口处；
    对所述目标货道中的商品进行盘点。
PCT/CN2021/114166 2020-08-31 2021-08-24 商品盘点方法和自动售货机 WO2022042507A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010899381.7A CN111986383B (zh) 2020-08-31 2020-08-31 商品盘点方法和自动售货机
CN202010899381.7 2020-08-31

Publications (1)

Publication Number Publication Date
WO2022042507A1 true WO2022042507A1 (zh) 2022-03-03

Family

ID=73446793

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/114166 WO2022042507A1 (zh) 2020-08-31 2021-08-24 商品盘点方法和自动售货机

Country Status (2)

Country Link
CN (1) CN111986383B (zh)
WO (1) WO2022042507A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115482533A (zh) * 2022-09-19 2022-12-16 北京百度网讯科技有限公司 陈列柜的拆分方法、装置、电子设备及存储介质

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111986383B (zh) * 2020-08-31 2022-03-25 威海新北洋数码科技有限公司 商品盘点方法和自动售货机
CN112773138A (zh) * 2021-02-04 2021-05-11 深圳市一方盒商业管理有限责任公司 商品推进装置及商品管理方法
CN113487788A (zh) * 2021-07-26 2021-10-08 美佳德科技（无锡）有限公司 一体化自动贩卖机管理系统及其装置

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2015076949A1 (en) * 2013-11-19 2015-05-28 Pantry Retail, Inc. Vending kit and method
CN107730168A (zh) * 2017-09-28 2018-02-23 中南大学 一种基于图像识别的自动售货机自动库存管理系统及方法
CN107993343A (zh) * 2017-12-11 2018-05-04 深圳市满心科技有限公司 自动售货机及盘点货物的方法以及存储介质
CN108629659A (zh) * 2018-04-27 2018-10-09 北京无人店科技有限公司 利用视觉测量进行盘点的无人售货系统
CN109427134A (zh) * 2017-08-31 2019-03-05 山东新北洋信息技术股份有限公司 一种自动售货机及存货量盘点方法
CN110766860A (zh) * 2018-07-27 2020-02-07 威海新北洋数码科技有限公司 一种自动售货机
CN210348649U (zh) * 2019-09-05 2020-04-17 上海了物网络科技有限公司 适用于智能售货的储物柜
WO2020082990A1 (zh) * 2018-10-23 2020-04-30 山东新北洋信息技术股份有限公司 自动售货机及其货物检测方法、电子设备和存储介质
CN111539468A (zh) * 2020-04-17 2020-08-14 深圳市智莱科技股份有限公司 一种自动售卖设备的货物清点方法
CN111986383A (zh) * 2020-08-31 2020-11-24 威海新北洋数码科技有限公司 商品盘点方法和自动售货机

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4049100B2 (ja) * 2004-01-08 2008-02-20 富士電機リテイルシステムズ株式会社 自動販売機、自動販売機の商品搬送方法
CN210155761U (zh) * 2019-09-20 2020-03-17 威海新北洋数码科技有限公司 具有提示组件的自动售货机

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2015076949A1 (en) * 2013-11-19 2015-05-28 Pantry Retail, Inc. Vending kit and method
CN109427134A (zh) * 2017-08-31 2019-03-05 山东新北洋信息技术股份有限公司 一种自动售货机及存货量盘点方法
CN107730168A (zh) * 2017-09-28 2018-02-23 中南大学 一种基于图像识别的自动售货机自动库存管理系统及方法
CN107993343A (zh) * 2017-12-11 2018-05-04 深圳市满心科技有限公司 自动售货机及盘点货物的方法以及存储介质
CN108629659A (zh) * 2018-04-27 2018-10-09 北京无人店科技有限公司 利用视觉测量进行盘点的无人售货系统
CN110766860A (zh) * 2018-07-27 2020-02-07 威海新北洋数码科技有限公司 一种自动售货机
WO2020082990A1 (zh) * 2018-10-23 2020-04-30 山东新北洋信息技术股份有限公司 自动售货机及其货物检测方法、电子设备和存储介质
CN210348649U (zh) * 2019-09-05 2020-04-17 上海了物网络科技有限公司 适用于智能售货的储物柜
CN111539468A (zh) * 2020-04-17 2020-08-14 深圳市智莱科技股份有限公司 一种自动售卖设备的货物清点方法
CN111986383A (zh) * 2020-08-31 2020-11-24 威海新北洋数码科技有限公司 商品盘点方法和自动售货机

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115482533A (zh) * 2022-09-19 2022-12-16 北京百度网讯科技有限公司 陈列柜的拆分方法、装置、电子设备及存储介质

Also Published As

Publication number Publication date
CN111986383A (zh) 2020-11-24
CN111986383B (zh) 2022-03-25

Similar Documents

Publication Publication Date Title
WO2022042507A1 (zh) 商品盘点方法和自动售货机
US7930064B2 (en) Automated drug discrimination during dispensing
US8706293B2 (en) Vending machine with automated detection of product position
US8457384B2 (en) Universal counting and measurement system
US8473374B2 (en) Method and apparatus for monitoring inventory
US7086560B2 (en) Vending machine
US9103816B2 (en) Blood sample analyzing apparatus, blood sample coagulation determining apparatus, blood sample coagulation determining method and computer program product
US9406041B2 (en) Methods and systems for automated monitoring and managing of inventory
US20050216120A1 (en) Automatic vending machine and method
WO2019201261A1 (zh) 货物数量检测方法及自动售货机
CN108629659B (zh) 利用视觉测量进行盘点的无人售货系统
CN111539468B (zh) 一种自动售卖设备的货物清点方法
US7628064B1 (en) Ultrasonic height and volume measuring instrument for laboratory test containers
DE19509631A1 (de) Verfahren und Einrichtung zum Erkennen und Sortieren von Getränkebehältern, insbesondere Flaschen und Flaschenkästen
WO2020020338A1 (zh) 一种自动售货机
JP5161477B2 (ja) 調剤装置のカウント装置
CN109858602A (zh) 一种商品数量检测装置、方法及自动售货机
US9539179B2 (en) Interactive kiosk for counting, labeling, and shipping of recalled medication capsules
CN111223239A (zh) 侧面出货自动售货机及出货结构
CN210924752U (zh) 一种侧面出货自动售货机及出货结构
CN210924745U (zh) 基于图像检测的自动售货机
CN113538784B (zh) 智能货柜以及物品识别方法
JP5359635B2 (ja) 賞味期限読取装置
JP2001258993A (ja) 物品収納装置
JP2008259530A (ja) 調剤装置のシート取り扱い方法及びその装置

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: 21860342

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: 21860342

Country of ref document: EP

Kind code of ref document: A1