WO2021170126A1 - 货品识别方法、货品识别系统及电子设备 - Google Patents

货品识别方法、货品识别系统及电子设备 Download PDF

Info

Publication number
WO2021170126A1
WO2021170126A1 PCT/CN2021/078316 CN2021078316W WO2021170126A1 WO 2021170126 A1 WO2021170126 A1 WO 2021170126A1 CN 2021078316 W CN2021078316 W CN 2021078316W WO 2021170126 A1 WO2021170126 A1 WO 2021170126A1
Authority
WO
WIPO (PCT)
Prior art keywords
goods
pallet
product
value
database
Prior art date
Application number
PCT/CN2021/078316
Other languages
English (en)
French (fr)
Inventor
冯立男
张一玫
Original Assignee
上海追月科技有限公司
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 上海追月科技有限公司 filed Critical 上海追月科技有限公司
Publication of WO2021170126A1 publication Critical patent/WO2021170126A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01G—WEIGHING
    • G01G19/00—Weighing apparatus or methods adapted for special purposes not provided for in the preceding groups
    • G01G19/40—Weighing apparatus or methods adapted for special purposes not provided for in the preceding groups with provisions for indicating, recording, or computing price or other quantities dependent on the weight
    • G01G19/413—Weighing apparatus or methods adapted for special purposes not provided for in the preceding groups with provisions for indicating, recording, or computing price or other quantities dependent on the weight using electromechanical or electronic computing means
    • G01G19/414—Weighing apparatus or methods adapted for special purposes not provided for in the preceding groups with provisions for indicating, recording, or computing price or other quantities dependent on the weight using electromechanical or electronic computing means using electronic computing means only
    • G01G19/4144—Weighing apparatus or methods adapted for special purposes not provided for in the preceding groups with provisions for indicating, recording, or computing price or other quantities dependent on the weight using electromechanical or electronic computing means using electronic computing means only for controlling weight of goods in commercial establishments, e.g. supermarket, P.O.S. systems
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01G—WEIGHING
    • G01G19/00—Weighing apparatus or methods adapted for special purposes not provided for in the preceding groups
    • G01G19/40—Weighing apparatus or methods adapted for special purposes not provided for in the preceding groups with provisions for indicating, recording, or computing price or other quantities dependent on the weight
    • G01G19/413—Weighing apparatus or methods adapted for special purposes not provided for in the preceding groups with provisions for indicating, recording, or computing price or other quantities dependent on the weight using electromechanical or electronic computing means
    • G01G19/414—Weighing apparatus or methods adapted for special purposes not provided for in the preceding groups with provisions for indicating, recording, or computing price or other quantities dependent on the weight using electromechanical or electronic computing means using electronic computing means only
    • G01G19/415—Weighing apparatus or methods adapted for special purposes not provided for in the preceding groups with provisions for indicating, recording, or computing price or other quantities dependent on the weight using electromechanical or electronic computing means using electronic computing means only combined with recording means
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/901—Indexing; Data structures therefor; Storage structures
    • 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

Definitions

  • the invention relates to a product perception technology used in the retail industry, in particular to a product identification method, a product identification system and electronic equipment.
  • the unmanned supermarket project In the traditional way of retail shopping, each supermarket or convenience store needs to have dedicated sales personnel and collection personnel, and the labor cost is relatively high. With the development of electronic payment technology, identity perception technology and cloud computing technology, the unmanned supermarket project has high technical feasibility. In the unmanned supermarket project, a basic problem that needs to be solved is the judgment and recording of the user's purchase of goods. Specifically, the server needs to accurately determine the type of goods that the user takes from the shelf, the quantity and unit price of the goods taken, etc. , In order to automatically realize settlement for users.
  • the same kind of goods are placed on the same shelf or pallet as much as possible. If a user removes a certain item from one shelf or a pallet and places it on another shelf or pallet, it will cause wrong picking.
  • the user For the problem of misplacement, for some solutions in the prior art, during the shopping process, the user must put the picked goods back to the original position when they put them back on the shelf.
  • the computer cannot accurately identify the type and quantity of the returned goods, and cannot update the user’s shopping database in a timely and accurate manner.
  • the user returns the goods to the wrong position on the shelf. It is necessary to pay for the product, causing identification errors and affecting user experience.
  • the purpose of the present invention is to provide a product perception technology based on micro-weight monitoring to solve the problem that the prior art cannot accurately distinguish the difference in details of the same type of goods, cannot accurately identify the identity of the changed goods, and is prone to mistaking and misplacement. problem.
  • the present invention provides a method for identifying goods, including the following steps: building a goods database, the goods database including at least one accurate weight value of the goods and the accurate weight value of at least one pallet; the accurate weight value is accurate to 10 x grams, x is an integer, -3 ⁇ x ⁇ 0; collect the accurate weight change value of each pallet on a shelf in real time; when the accurate weight change value of a pallet is a non-zero value, according to the accurate weight change value Identify whether any goods are taken or put back, obtain the goods number corresponding to the goods; and store the real-time accurate weight sensing value of the pallet in the goods database to replace the accurate weight value of the pallet in the goods database.
  • the goods identification method further includes the following steps: identifying the identity information of each user; generating the user’s shopping database according to the user’s identity information; The identity of the user who walked or put back the goods.
  • the method further includes the following steps: calling the product information of the product from the product database according to the product number; and when a product is taken from a pallet, The product information of the product is recorded in the shopping database of the user who took the product; when the product is put back on a pallet, the product information of the product is deleted from the shopping database of the user who put the product back.
  • the accurate weight change value of the pallet is less than 0; if so, it is determined that goods have been taken from the pallet; if If no, it is determined that there are goods or items to be returned to the pallet.
  • the step of obtaining the product number corresponding to the product it is determined whether the absolute value of the precise weight change value is equal to the precise weight value of a product in the product database; if so, the product in the product database is determined It is the goods that have been taken away or put back, and the product number is obtained.
  • the step of obtaining the item number corresponding to the item includes the following steps: when the accurate weight value of a tray is greater than zero, determining the identity of the user who put the item back; calling the shopping database of the user who put the item back; determining Whether the absolute value of the accurate weight change value is the same as the accurate weight value of an item in the user's shopping database; if so, determine that the item is a returned item, and obtain the item number of the item.
  • the step of constructing the goods database specifically includes the following steps: initializing the goods database, entering the shelf number, pallet number, and the accurate weight value of each pallet; in the initial state, the accurate weight value of each pallet in the goods database It is the accurate weight sensing value of the pallet when there is no goods; enter the goods information of a goods into the goods database, and the goods information includes the goods number, the type of the goods, the standard weight value of the goods of the type, and the price of the goods ; After an item is placed on a pallet of a shelf, the accurate weight change value of each pallet is collected in real time; when the accurate weight change value of a pallet is non-zero, the pallet’s value is recorded in the goods database The accurate weight change value is the accurate weight value of the goods; and the real-time accurate weight sensing value of the pallet is entered into the goods database to replace the accurate weight value of the pallet in the goods database.
  • the accurate weight change value of the pallet in the goods database is recorded as the accurate weight value of the goods.
  • the accurate weight change value and the accurate weight change value are calculated.
  • the difference between the standard weight value of a type of goods; when the ratio of the absolute value of the difference to the standard weight value of the type of goods is less than a preset threshold, the accurate weight change value of the pallet is recorded in the goods database. The precise weight value of the goods.
  • the step of collecting the accurate weight change value of each pallet on a shelf in real time specifically includes the following steps: collecting the real-time accurate weight sensing value of each pallet; and calculating the accurate weight change value of each pallet on the shelf , Which is the difference between the real-time accurate weight sensing value of each pallet and the accurate weight value of the pallet in the goods database.
  • the weight sensing value of a pallet when the weight sensing value of a pallet changes, it is determined whether the weight sensor can collect accurate and stable sensing data. If not, it is determined that the pallet is in an abnormal state. Selectively send out an alarm signal.
  • the step of constructing the goods database further includes the following steps: setting at least one shelf, each shelf including a plurality of independent trays; and setting a micro-weight sensor under each tray, the weight of the micro-weight sensor
  • the sensing value is accurate to 10 x grams, x is an integer, -3 ⁇ x ⁇ 0; place at least one product in at least one pallet, and only one product at a time.
  • the present invention also provides an electronic device, including a memory and a processor; the memory is used to store executable program code; the processor is connected to the memory, and the executable program code is read by the processor to run and The computer program corresponding to the executable program code executes at least one step in the method for identifying goods.
  • the present invention also provides a product identification system, which includes the electronic equipment described above, at least one shelf and at least one micro-weight sensor; each shelf includes a plurality of independent trays; at least one micro-weight sensor is arranged under each tray, And connected to the electronic device.
  • the micro-weight sensor is used to obtain the accurate weight value of at least one product and/or the accurate weight value of the tray; the accurate weight value is accurate to 10 x grams, x is an integer, and -3 ⁇ x ⁇ 0.
  • the beneficial effect of the present invention is to provide a product identification method, a product identification system and electronic equipment.
  • a high-precision micro-weight sensor in the pallet of the shelf, the accurate weight change value of each pallet on the shelf is monitored in real time. (Accurate to 10 x grams, x is an integer, -3 ⁇ x ⁇ 0), according to the real-time changes of the accurate weight value of all shelves on the shelf, it can be inferred which goods have been taken or put back, not only can judge The type of the goods that have been taken away or put back, and the identity of the goods can be accurately determined and the product number can be obtained.
  • the computer can accurately identify the item number corresponding to the item that was removed or put back, and then accurately update the user Shopping information. Even if the user picks up the wrong item, the user’s shopping record will not go wrong, even if the item is placed in the wrong place, it will not affect the purchase process of the next user, which fundamentally solves the problem of picking up the wrong item. The misplacement problem improves the user experience.
  • Figure 1 is a top view of an unmanned supermarket according to an embodiment of the present invention.
  • Figure 2 is a schematic diagram of the overall structure of the shelf according to an embodiment of the present invention.
  • FIG. 3 is a schematic diagram of the structure of the tray and the shelf according to the embodiment of the present invention.
  • FIG. 4 is a flowchart of a method for identifying goods according to an embodiment of the present invention.
  • FIG. 5 is a flowchart of the steps of constructing a goods database according to an embodiment of the present invention.
  • FIG. 6 is a flowchart of the first collecting step of the tray change value according to the embodiment of the present invention.
  • FIG. 7 is a flowchart of the second collecting step of the tray change value according to the embodiment of the present invention.
  • FIG. 8 is a schematic diagram of the structure of the goods identification system according to an embodiment of the present invention.
  • the part When some part is described as being “on” another part, the part may be directly placed on the other part; there may also be an intermediate part on which the part is placed, And the middle part is placed on another part.
  • a component When a component is described as “installed to” or “connected to” another component, both can be understood as directly “installed” or “connected”, or a component is indirectly “mounted to” or “connected to” through an intermediate component “Another component.
  • This embodiment relates to a goods identification system, which is a part of an unmanned vending system used in an unmanned supermarket or an unmanned convenience store.
  • the unmanned vending system includes a closed space 1 in which a plurality of shelves 2 are provided, and each shelf 2 includes a bracket 3 and a plurality of detachable mounts on the bracket 3.
  • the shelf board 4, a plurality of shelf boards 4 are parallel to each other at different heights or flush with each other at the same height.
  • Each shelf 4 is provided with a pallet 5 or a plurality of pallets 5 arranged side by side, each pallet 5 is placed with at least one kind of goods, and the number of each kind of goods is one or more.
  • the types of goods on the same pallet 5 are the same or different.
  • only one kind of goods is placed on the same pallet 5.
  • the goods identification system 100 of the present invention includes an electronic device 10, at least one cargo rack 2, and at least one micro-weight sensor 6.
  • At least one micro-weight sensor 6 is arranged between each tray 5 and the shelf 4, preferably a weight sensor approximately in the shape of a rectangular parallelepiped.
  • the lower surface of one end is connected to a shelf 4, and the upper surface of the other end is connected to a
  • each micro-weight sensor 6 is connected to the electronic device 10, and the total accurate weight value of the tray 5 above it and all the goods placed on the upper surface of the tray 5 is obtained in real time and the accurate weight value is transmitted to the electronic device 10.
  • the electronic device 10 is preferably a computer, and a product database is stored in the memory of the product database.
  • the product information of at least one product is stored in the product database; the product information includes the product category, product price, and standard weight value of each type of product. Further, the product information also includes the product number of each product, and the information of each product can be searched according to the product number.
  • the goods database also stores the number m of each pallet, the precise weight value Pm0 of each pallet in an empty state, the type of goods placed on each pallet, the precise weight value of each product, and each product The placement and so on. Goods of the same type have the same standard weight value, but due to individual differences, different goods of the same type have different accurate weight values.
  • the accuracy of the sensor is as long as it can be It is enough to judge the individual shipment, and the operator of the convenience store can choose the appropriate sensor accuracy according to the total number of goods.
  • goods of the same type have the same standard weight value, but due to individual differences, there is usually a certain deviation between the accurate weight value of a product and its standard weight value, and the deviation is generally less than or equal to 5%.
  • the deviation is generally less than or equal to 5%.
  • the weight accuracy of the micro-weight sensor is only accurate to 1 gram, several different products of the same type are likely to have different weight values, and the possibility that the exact weight values of two or more products are exactly the same will be further reduced.
  • the standard weight value of a product is 100 grams, and the reasonable deviation range of its weight value is plus or minus 5%
  • the true weight value of each product in this category may be any one between 95 and 105 grams.
  • the probability that the exact weight values of two items are exactly the same is 1/11, or 9.1%.
  • the weight accuracy of the micro-weight sensor is accurate to 0.1 gram, the probability that the exact weight values of two items are exactly the same is 0.91%.
  • the weight accuracy of the micro-weight sensor is accurate to 0.01 gram, there will be two The probability that the exact weight value of each item is exactly the same is 0.091%.
  • the total number of goods placed on the shelf in an ordinary convenience store is 300-3000 pieces, and the number of goods of the same type placed on the same tray is generally only 5-30, so there are two convenience stores. The possibility that the exact weight of the product is exactly the same is very small and can be ignored.
  • micro-weight sensors in the prior art have a weight accuracy of 10 -4 grams.
  • the computer can ignore the impact of small probability events and think that each item has a unique accurate weight value. Therefore, in this embodiment, the precise weight value of the goods can be used as a parameter specific to each item, and the precise weight value can also be called micro-weight value, micro-weight parameter or micro-weight attribute.
  • the electronic device of the present invention uses the accurate weight value of each product as the basis for judging the product taking and placing status, and accurately determines the product number of the removed or returned product according to the weight change of each pallet on the shelf, and then Call the product information such as the price and type of the product from the product database. After determining the identity of the user who fetches and releases the goods, the user's shopping database is updated in real time.
  • the micro-weight sensor under the pallet where the product is located collects the accurate weight value of the pallet in real time, compares it with the accurate weight value of the pallet at the previous moment in the goods database, and calculates it. The difference between the two to obtain the accurate weight change value of the pallet. Since the serial number of each pallet and the type of the goods placed on each pallet are pre-stored in the goods database, the electronic device 10 can obtain the possible types of the goods to be taken according to the serial number of the pallet whose weight changes. Since the standard weight value of each type of product and the precise weight value of each product are pre-stored in the product database, the electronic device 10 can determine the product category and product number of the removed product based on the accurate weight change value of the pallet .
  • each product corresponds to a micro-weight parameter, three thousand products will have three thousand accurate weight values. Even if there are many different types of goods placed in the same pallet, the electronic device can accurately recognize that they are taken The item number corresponding to the item.
  • the electronic device 10 stores the new accurate weight value of the pallet in the product database to replace the original accurate weight value of the pallet, and complete the update of the product database.
  • the electronic device 10 writes the product information of the removed product into the shopping database of the user who has removed the product, so as to update the user's shopping record in real time.
  • the electronic device can also accurately identify the product number corresponding to the returned product, and the data processing flow is similar to the previous process.
  • the electronic device stores the new accurate weight value of the pallet in the goods database to replace the original accurate weight value of the pallet to complete the update of the goods database.
  • the electronic device 10 deletes the product information of the returned product from the shopping database of the user who returned the product to update the user's shopping record in real time.
  • the present invention provides a method for identifying goods, which includes the following steps S1-S12.
  • the shelf setting step is to set at least one shelf, and each shelf includes a plurality of independent trays.
  • the sensor setting step is to install at least one micro-weight sensor under each tray to monitor the change in the weight of the tray in real time.
  • the precise weight value measured by the micro-weight sensor is accurate to 10 x grams, x is an integer, -3 ⁇ x ⁇ 0.
  • the types of different goods on the same pallet can be the same or different.
  • the staff placed each product in turn, one piece at a time, so that the electronic device can collect the precise weight value of each product, that is, the micro-weight parameter.
  • the staff before placing the goods, sets the type of goods that should be placed on each pallet in the electronic device (computer), and at the same time enters a number of goods information for each type of goods, such as the standard weight value of the goods, Variety of goods, place of origin and price of goods, etc.
  • a barcode or two-dimensional code is provided on the outer packaging of each product, and several product information of the product is recorded.
  • a scanning device such as a barcode reader or a QR code reader
  • a two-dimensional code to obtain and transmit the product information of the product to the electronic device (computer) product database.
  • the staff puts the same kind of goods to a certain pallet in turn, the electronic device obtains the number of the pallet whose weight has changed, and sets the type of goods that should be placed on the pallet to be the goods recorded by the previous scan code type.
  • the step of constructing a goods database constructing a goods database, in which the goods information of at least one goods is stored; the goods information includes the type of goods, the price of the goods and the standard weight value of each type of goods. Further, the product information also includes a product number, which corresponds to each product one-to-one, so that it is convenient for the computer to find the product.
  • the pallet number m of each pallet, the precise weight value Pm0 of each pallet in an empty state, the precise weight value of each product, and the placement position of each product are also recorded in the product database.
  • the precise weight value is accurate to 10 x grams, x is an integer, and -3 ⁇ x ⁇ 0; goods of the same type have the same standard weight value, but due to individual differences, different goods of the same type have different precise weight values.
  • the goods database construction step S4 and the goods placement step S3 are carried out synchronously.
  • the accurate weight value of the goods is collected once the goods are placed; during the goods placement process, each time a product is placed, the micro-weight sensor of the shelf obtains the pallet
  • the real-time weight change value of the product is transmitted to the electronic device (computer).
  • the electronic device will obtain a weight-sensing change value in real time and record it as the accurate weight value of the product, which is stored in the product database.
  • the computer generates the weight based on the shelf
  • the number of changes can be recorded for the number of each product. For each change, the counter is incremented by one. According to the number of the pallet whose weight has changed, the position of the pallet for each product can be obtained.
  • the computer can accurately identify the item numbers of the items that have been taken or put back, so that there will be no mistakes in picking and placing the goods and causing them to shop.
  • the goods database construction step S4 specifically includes the following steps S41-S46.
  • S41 The step of initializing the goods database, initializing the goods database, entering the shelf number, pallet number, and the precise weight value of each pallet; at the same time, the type of goods placed on each pallet of each shelf can be preset in the goods database.
  • the accurate weight value of each pallet in the goods database is the accurate weight sensor value of the pallet in the state of no goods.
  • the entry step of the category of goods, the external QR code or barcode of each product stores the product information, each product must be scanned once before being placed on the shelf, and the product of each product is entered by scanning the code.
  • the information is sent to the product database, and the product information includes the category of the product, the standard weight value of the category of the product, and the price of the product, and may also include the product number.
  • the product information includes the category of the product, the standard weight value of the category of the product, and the price of the product, and may also include the product number.
  • S43. In the first collecting step of the pallet change value, after a product is placed on a pallet of a shelf, the real-time accurate weight change value Qmi of each pallet is collected in real time. m is the pallet number, i is the number of weight changes of each pallet during the placement of the goods, and the number of the placed goods can be calculated according to the number of weight changes i. S44.
  • the first judging step of the pallet change value is to judge whether the accurate weight change value Qmi of all pallets is all 0; if yes, return to the goods category entry step S42; if not, execute the next step. S45. Goods weight recording step.
  • the accurate weight change value Qmi of a pallet is a non-zero value, calculate the difference between the accurate weight change value Qmi and the standard weight value A of the type of goods (Qmi-A); when the difference is When the ratio of the absolute value of the value
  • the preset threshold value is a certain deviation ratio of the standard weight value A of the type of goods, such as 3%-5% of A. It is a manual setting method. The value can be changed according to the actual situation to prevent entry errors of non-type goods. Prevent serious quality problems of this kind of goods.
  • the first update step of the goods database is to enter the real-time accurate weight sensing value of the pallet into the goods database to replace the accurate weight value of the pallet in the goods database; return to the goods category entry step S42.
  • the product information of each product can be accurately entered into the product database, in particular, it can be realized that the accurate weight value Qmi of each product corresponds to a product number, and one product number corresponds to a product price.
  • the goods number corresponding to the goods that are taken or put back can be identified.
  • step S46 after the accurate weight value data of the pallet in the goods database is updated, return to the goods category entry step S42, and repeat steps S42-S46 once for each item placed.
  • This entry method can accurately enter the precise weight value of a product corresponding to each product number.
  • the staff before placing each type of goods, the staff first scans the barcode or two-dimensional code of a goods with a code scanning device, and obtains and transmits the goods information of that type of goods to the electronic device (computer ), stored in the goods database. After that, the staff placed multiple items of this type in sequence, and the shelf obtained the accurate weight value of each item in real time and stored it in the item database.
  • step S42 the product information of a product is entered into the product database.
  • the product information includes the product category, the standard weight value of the product, and the price of the product, but does not include the product number of the product.
  • step S43 multiple products of this type are placed on a pallet of a shelf in sequence, and the shelf records the product number and precise weight value of each product according to each change in the weight value of the pallet, and other products of the same category.
  • the product information is the same.
  • This entry method can sequentially enter each precise weight value of the same type of goods through the type of goods, which improves the efficiency of entering the precise weight value of the goods, and the prices of the goods of the same type are the same, and there will be no problem of settlement errors.
  • the first collecting step S43 of the tray change value specifically includes the following steps S431-S432. S431.
  • the first collection step of pallet weight after a product is placed on a pallet of a shelf, the real-time accurate weight sensing value Pmi of each pallet is collected, m is the pallet number, and i is each item during the placement process. The number of changes in the weight of the pallet. S432.
  • the type of the newly added product can be obtained first, and then based on the type of the newly added product, the difference of each precise weight value of multiple newly added products can be obtained continuously Qmi is used as the accurate weight value of each new product to be entered into the product database, which can quickly enter the accurate weight value information of the new product, and will not affect the settlement of the price of the product.
  • the goods database construction step S4 can also be a manual entry method.
  • the staff uses a high-precision electronic scale to sequentially measure and record the precise weight value of each product; at the same time, the product information can be manually entered, such as the product. Number, product category and product price, etc.; you can also enter the shelf number and pallet number of each product.
  • the accurate weight change value of each pallet on a shelf is collected in real time.
  • the second collecting step S5 of the tray change value specifically includes the following steps S51-S52.
  • the second collecting step of pallet weight is to collect real-time accurate weight sensing value of each pallet on a shelf in real time.
  • the real-time accurate weight sensing value Pni, n is the pallet number
  • i is the number of changes in the weight of each pallet during the placement of the goods.
  • the accurate weight change value Qni of each pallet can be obtained in turn.
  • the accurate weight change value Qni of the pallet according to the shelf With the accurate weight value Qmi of the goods in the goods database, it is possible to identify the picked or The item number of the item to be returned.
  • This method completely relies on the precise weight value (microweight parameter) of each item to identify the item.
  • the problem of shopping record errors and shopping settlement errors fundamentally solves the problem that users’ shopping records are wrong due to misplaced items during shopping. If the gravity of the pallet and the goods changes rapidly in a very short time, the sensing data of the weight sensor will also change quickly, and the accurate weight value cannot be collected. Only in a stable state, the weight sensor can collect accurate weight sensing value.
  • step S51 when the weight sensing value of a tray changes, it is judged whether the weight sensor can collect accurate and stable sensing data, if yes, go to step S52; if not, it is judged that the tray is in an abnormal state, which can also be called It is an unstable abnormal state; the reason may be that the user's body or belongings are leaning on the shelf, it may be that the user drags or presses the goods on the tray, or it may be that insects are crawling on the tray, and so on.
  • an alarm signal is issued, and after waiting for a certain period of time (a preset period of time, such as 1 to 3 seconds), if the abnormal status is eliminated, step S51 is executed again.
  • the alarm signal can be the sound of a buzzer, an abnormal reminder or an abnormal information record on the display screen to remind the manager or the user.
  • the second judgment step of the pallet change value is to judge whether the accurate weight change values Qni of all the pallets are all 0; if yes, return to the second collecting step S5 of the pallet change value; if not, execute the next step.
  • the precise weight change value of a pallet is the precise weight value of the goods that are removed or put back on the pallet. If the precise weight change values of all pallets are all 0, it means that no goods have been taken or put back. Return to step S5 to prepare for the next collection; if the precise weight change value of a certain tray is a non-zero value, it can be judged that the goods or foreign objects are removed or put back. The foreign objects are those that do not belong to the convenience store. Items in stock. S7.
  • the step of judging the pick-and-place state it is judged whether the accurate weight change value Qni of the pallet is less than 0, if yes, it is judged that there are goods removed from the pallet, and if not, it is judged that there are goods or objects are put back to the pallet.
  • step S8 it is determined whether the absolute value of the precise weight change value
  • the item database includes the pallet number of the pallet where each item is located; items of the same type are placed on the same pallet.
  • the goods judging step S8 includes the following two situations: when the accurate weight value of a pallet is a negative value, call the accurate weight value of each item on the pallet from the goods database according to its pallet number; determine the accurate weight Whether the absolute value of the change value is equal to the exact weight value of a product on the pallet, if so, determine that the product is the removed product, and obtain the product number of the product.
  • the second updating step of the goods database storing the real-time accurate weight sensing value of the pallet in the goods database to replace the accurate weight value of the pallet in the goods database, and returning to the second collecting step S5 of the pallet change value. Ready to monitor the next change of goods on each shelf. If the weight sensing value of a pallet changes steadily, the real-time accurate weight sensing value of the pallet in the goods database does not need to be updated; if the weight sensing value of a pallet changes steadily, no matter whether its change value corresponds to a certain product in the goods database Regardless of whether the item being taken or put back is the original item of the convenience store, the accurate weight sensing value of the corresponding pallet stored in the item database needs to be updated in real time.
  • the abnormal state in this embodiment refers to a situation where the real-time weight sensing value of the tray changes unexpectedly, including an unstable abnormal state and a stable abnormal state.
  • the weight sensor cannot obtain a stable real-time weight sensing value.
  • the cause may be that someone deliberately exerts upward force on a product, someone drags the product on the tray, someone leans or presses the tray, or insects. Stay or crawl on the pallet, etc.
  • the unstable abnormal state lasts for a relatively short time. Therefore, after waiting for a period of time, the weight sensing value can be collected again after the abnormal state is eliminated.
  • the weight sensor can obtain a stable real-time weight sensing value, but at this time, the difference in the weight change of the pallet cannot form a match with any product in the product database.
  • the electronic device cannot confirm the identity of the product, and the The item is deleted from the user's shopping database.
  • the cause of the stable abnormal state may be that the user puts an item that does not belong to the convenience store on the tray, or puts a part of the food that has been eaten back on the tray.
  • the stable abnormal state lasts for a relatively long time.
  • the real-time weight sensing value of the pallet can be stored in the goods database to replace the accurate weight value of the pallet in the goods database. Regardless of whether the abnormal state is eliminated, when other goods in the tray are picked and placed by the user, the electronic device can still record the correct shopping behavior, and the experience of other users will not be affected.
  • the goods identification method further includes the following steps S21-S24.
  • S21 the user identity recognition step, to identify the identity information of each user.
  • the aforementioned shelves and trays are arranged in a closed space 1 to turn the closed space into an unmanned convenience store.
  • a gate is set at the entrance and exit of the enclosed space 1, and an identity recognition device is set on the gate.
  • a two-dimensional code reading and writing device is preferred to read the identity two-dimensional code in the user's mobile phone APP to identify the user's identity.
  • the user can bind an electronic payment account in the mobile APP, and authorize it to the operator of the convenience store, allowing the operator to deduct the payment.
  • a shopping database generating step generating a shopping database of the user according to the user's identity information.
  • the shopping database is the shopping information corresponding to each user, including the user's identity information and the goods information corresponding to the goods taken away by the user, including the type, quantity and price of the goods taken away, so as to facilitate the calculation of convenience store operators The amount the user needs to pay in order to settle for the customer.
  • the shopping user judging step when the goods are taken or put back, judge the identity of the user who took or put the goods back.
  • the weight of a pallet changes and the user's hand interacts with the shelf at the same time, it is determined that the user has taken or put back the goods.
  • the interaction between the user's hand and the shelf means that the computer judges that a user's hand coordinates coincide with the pallet whose weight changes, or that a user's hand appears near the pallet whose weight changes, and so on.
  • the shopping database update step after each product judgment step S8, the shopping database of the corresponding user is updated in real time.
  • the product information of the product is called from the product database according to the product number, and the product information of the product is recorded in the shopping database of the user who took the product; the product information includes the product The type, the standard weight value of the type of goods and the price of the goods, etc.; when the goods are put back on the shelf, the goods information of the goods will be deleted from the shopping database of the user who put the goods back.
  • the shopping database update step updates the shopping record of each user in real time according to the real-time shopping behavior of each user, so that the actual shopping situation of each user is completely consistent with the shopping record in the electronic shopping cart for the user to settle.
  • the probability that two goods with exactly the same weight value may appear in the entire convenience store will also be higher, and the computer recognition error rate will be increased.
  • the existing weight sensor can collect high-precision weight values, the higher the accuracy of the sensor, the higher the manufacturing requirements, the higher the cost, and the more expensive the price.
  • a large number of weight sensors are required in the present invention, which will cause problems. Operators of human convenience stores bring greater cost pressure.
  • each item in a convenience store has a unique accurate weight value (microweight parameter), and the computer can effectively distinguish each item based on this.
  • the exact weight value of two or more different types of goods is the same.
  • the accuracy of the weight sensor is required to be high, and the accurate weight value is accurate to 10 -2 grams or 10 -3 grams to ensure that each product has a unique accurate weight value. Since the overall solution requires the use of a large number of weight sensors, the hardware cost of high-precision sensors is relatively high, which will cause the cost of the entire solution to be too high and affect the promotion and application of this patent.
  • the electronic device 10 may first determine the identity of the user who puts back the goods, and may adopt the solution of scanning the code at the entrance guard to identify the user's identity and the solution of real-time monitoring of the user's location by a three-dimensional camera.
  • the hand coordinate set of a user with a known identity coincides with the coordinate set above a tray, and the weight sensing value of the tray changes at the same time, the identity of the user who took or put the goods back can be judged.
  • the electronic device in the S8 product determination step, when the accurate weight change value of a pallet is less than zero, the electronic device (computer) can determine that the product has been taken away; because it was previously placed on each pallet
  • the product category of is pre-stored in the product database, and the electronic device (computer) can determine the category of the product.
  • a pallet is placed on a pallet of goods of the same type, of the same brand, and approximately the same weight; when the weight of the pallet becomes lighter, the computer can determine the type of goods to be taken away; further, according to the accurate weight change value and the goods database By comparing the accurate weight value of each item on the pallet stored in the pallet, judge and record the item number of the item. Further, the identity of the user who took the goods is determined in the shopping user judgment step S23, and the goods information of the goods is recorded in the shopping database of the user who took the goods in the shopping database update step S24.
  • the S8 product judgment step when the precise weight change value of a tray is greater than zero, the S8 product judgment step includes the pick-and-place status judgment step, the shopping user judgment step, the shopping database call step, and the item Judgment steps.
  • the electronic device can determine that an item was put back, and then determine the identity of the user who put the item back, call the shopping database of the user who put the item back, and determine whether the absolute value of the accurate weight change value is the same as that of the user
  • the exact weight value of an item in the shopping database is the same; if it is, it is determined that the item is a returned item, and the item number of the item is obtained; if not, it is determined that the item is not the item that the user previously removed from the shelf, It is not the goods of this store, but may be the user’s personal belongings, such as gloves, mobile phones, etc.
  • the second update step S9 of the item database is performed after step S8, and the real-time accurate weight sensing value of the pallet is stored in the item database to replace the item database
  • the accurate weight value of the pallet is returned to the second collecting step S5 of the pallet change value to prepare to monitor the next product change of each shelf.
  • the S24 shopping database update step is executed to delete the goods information corresponding to the goods from the shopping database of the user.
  • the purchase amount of a user in a convenience store is generally only a few or a dozen, but at most dozens. These dozens of goods belong to different categories, including those with exactly the same microweight value.
  • the probability of two items will be less than one in ten million, which is completely negligible. Therefore, the accurate weight value of each item only needs to be accurate to 1 gram or 0.1 gram.
  • a low-precision and low-cost weight sensor can be used to effectively save operation and maintenance costs, further reduce the probability of computer errors, and improve recognition accuracy.
  • the user's shopping database may be updated one or more times.
  • a settlement step is also included to calculate the amount that the user needs to pay based on a user's shopping database.
  • the electronic device computer
  • the electronic device will think that the user has left, and calculate the user's consumption amount based on the user's shopping database
  • the convenience store operator can directly deduct the corresponding payment from the user’s electronic payment account. Users do not need to queue up for settlement, they can leave the market directly after shopping, and get a very good user experience.
  • the electronic device 10 described in this embodiment includes at least one memory 11 and at least one processor 12.
  • a storage medium is provided in the memory 11 for storing executable program codes.
  • the processor 12 is connected to the memory 11, and runs the computer program corresponding to the executable program code by reading the executable program code to execute steps S4-S9 and S21-S24 in the method for identifying goods. At least one step in.
  • the beneficial effect of the present invention is to provide a product identification system, a product identification method, storage medium and electronic equipment, which can monitor the accurate weight change value of each pallet on the shelf in real time, and determine the accurate weight change of all shelves on the shelf.
  • the phenomenon of the user taking the wrong goods by mistake will not affect the normal shopping settlement of the next user, and fundamentally solve the problem that the user’s shopping record is wrong due to the wrong goods being taken by the user during shopping.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Mathematical Physics (AREA)
  • General Physics & Mathematics (AREA)
  • Databases & Information Systems (AREA)
  • Software Systems (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Warehouses Or Storage Devices (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种货品识别方法、货品识别系统（100）及电子设备（10），通过实时监测货架（2）上的每一托盘（5）的精准重量变化值，可以判断被取走或被放回货品的货品编号，从而对于用户的购物信息判断更加精准，提升了用户体验。

Description

货品识别方法、货品识别系统及电子设备 技术领域
本发明涉及一种用于零售业的货品感知技术，具体地说，涉及一种货品识别方法、货品识别系统及电子设备。
背景技术
传统零售业的购物方式，每一家超市或便利店需要有专门的销售人员和收款人员，人力成本较高。随着电子支付技术、身份感知技术及云计算技术的发展，无人超市项目在技术上具备很高的可行性。在无人超市项目中，急需解决的一个基本问题就是用户选购货品的判断和记录问题，具体地说，服务器需要准确判定用户从货架上取走货品的种类、取走货品的数量及单价等，以便自动为用户实现结算。
现有技术中，同种类的货品尽量摆放在同一货架或托盘上，如果用户将某货品从一货架或一托盘取下后，将其放置到另一货架或托盘上，就会造成错拿错放的问题，对于现有技术的一些解决方案来说，在购物过程中，用户将已拾取的货品放回货架时必须放回原位。当用户在将已拾取的货品错放至其他位置时，计算机不能精准识别被放回货品的种类及数量，不能及时准确地更新用户的购物数据库，用户将货品放回至货架的错误位置后还要支付该商品费用，造成识别错误，影响用户体验。用户将错拿的货品乱放至错误的货架或托盘，进而导致该用户购物记录出错的问题，被业内称为错拿错放问题。
现有技术中，在货架上安装重量传感器只能判断被取走或被放回的货品属于哪一种类，而不能准确分辨同种类货品的细节差异，不能判断出被取走或被放回的货品是哪一个。因此，市场上急需一种能精准判断变动货品身份的技术，以解决错拿错放的问题。
发明内容
本发明的目的在于，提供一种基于微重量监测的货品感知技术，以解决现有技术不能 准确分辨同种类货品的细节差异、不能精准识别变动货品的身份、容易发生错拿错放现象的技术问题。
为实现上述目的，本发明提供一种货品识别方法，包括如下步骤：构建货品数据库，所述货品数据库包括至少一货品的精准重量值及至少一托盘的精准重量值；所述精准重量值精确到10 x克，x为整数，-3≤x≤0；实时采集一货架上的每一托盘的精准重量变化值；当一托盘的精准重量变化值为非零数值时，根据该精准重量变化值识别是否有货品被取走或被放回，获取该货品对应的货品编号；以及存储该托盘的实时精准重量感应值至所述货品数据库，以替换所述货品数据库中该托盘的精准重量值。
进一步地，所述货品识别方法还包括如下步骤：识别每一用户的身份信息；根据所述用户的身份信息生成该用户的购物数据库；以及当有货品被取走或被放回时，判断取走或放回该货品的用户的身份。
进一步地，在获取该货品对应的货品编号的步骤之后，还包括如下步骤：根据所述货品编号从所述货品数据库中调用该货品的货品信息；以及当有货品从一托盘被取走时，将该货品的货品信息记录在取走该货品的用户的购物数据库中；当有货品被放回至一托盘时，从放回该货品的用户的购物数据库中删除该货品的货品信息。
进一步地，在根据该精准重量变化值识别是否有货品被取走或被放回的步骤中，判断该托盘的精准重量变化值是否小于0；若是，判定有货品被从该托盘取走；若否，判定有货品被物品或放回至该托盘。
进一步地，在获取该货品对应的货品编号的步骤中，判断该精准重量变化值的绝对值与所述货品数据库中一货品的精准重量值是否相等；若是，判定所述货品数据库中的该货品即为被取走或被放回的货品，获取其货品编号。
进一步地，获取该货品对应的货品编号的步骤，包括如下步骤：当一托盘的精准重量值大于零时，判断放回该物品的用户的身份；调用放回该物品的用户的购物数据库；判断该精准重量变化值的绝对值是否与该用户的购物数据库中一货品的精准重量值相同；若是，判 定该物品为被放回的货品，获取该货品的货品编号。
进一步地，构建货品数据库的步骤，具体包括如下步骤：初始化货品数据库，录入货架编号、托盘编号及每一托盘的精准重量值；在初始状态下，所述货品数据库中每一托盘的精准重量值为该托盘在无货品状态下的精准重量感应值；录入一货品的货品信息至所述货品数据库，所述货品信息包括该货品的货品编号、货品种类、该种类货品的标准重量值以及货品价格；一货品被摆放至一货架的一托盘上之后，实时采集每一托盘的精准重量变化值；当一托盘的精准重量变化值为非零数值时，在所述货品数据库中记录该托盘的精准重量变化值为该货品的精准重量值；以及录入该托盘的实时精准重量感应值至所述货品数据库，以替换所述货品数据库中该托盘的精准重量值。
进一步地，在所述货品数据库中记录该托盘的精准重量变化值为该货品的精准重量值的步骤中，当一托盘的精准重量变化值为非零数值时，计算该精准重量变化值与该种类货品的标准重量值的差值；当该差值的绝对值与该种类货品的标准重量值的比值小于一预设阈值时，在所述货品数据库中记录该托盘的精准重量变化值为该货品的精准重量值。
进一步地，实时采集一货架上的每一托盘的精准重量变化值的步骤，具体包括如下步骤：采集每一托盘的实时精准重量感应值；以及计算该货架上的每一托盘的精准重量变化值，其为每一托盘的实时精准重量感应值与所述货品数据库中该托盘的精准重量值的差值。
进一步地，所述托盘重量第二采集步骤中，当一托盘的重量感应值发生变化时，判断重量传感器是否能采集到准确的、稳定的感应数据，若否，判断该托盘为异常状态，可选择地发出报警信号。
进一步地，在实时采集一货架上的每一托盘的精准重量变化值的步骤之后，判断一托盘精准重量变化值的绝对值是否与所述货品数据库中一货品的精准重量值相同；若否，判断该托盘为异常状态，可选择地发出报警信号。
进一步地，在构建货品数据库的步骤之前，还包括如下步骤：设置至少一货架，每一货架包括多个彼此独立的托盘；以及设置微重量传感器于每一托盘下方，所述微重量传感器 的重量感应值精确到10 x克，x为整数，-3≤x≤0；摆放至少一货品于至少一托盘内，每次仅摆放一个货品。
本发明还提供一种电子设备，包括存储器以及处理器；所述存储器用于存储可执行程序代码；所述处理器连接至所述存储器，通过读取所述可执行程序代码，来运行与所述可执行程序代码对应的计算机程序，以执行所述货品识别方法中的至少一步骤。
本发明还提供一种货品识别系统，包括前文所述电子设备、至少一货架以及至少一微重量传感器；每一货架包括多个彼此独立的托盘；至少一微重量传感器设于每一托盘下方，且连接至所述电子设备。其中，所述微重量传感器用以获取至少一货品的精准重量值和/或托盘的精准重量值；所述精准重量值精确到10 x克，x为整数，-3≤x≤0。
本发明的有益效果在于，提供一种货品识别方法、货品识别系统及电子设备，通过在货架的托盘内设置高精度的微重量传感器的方式，实时监测货架上的每一托盘的精准重量变化值(精确到10 x克，x为整数，-3≤x≤0)，根据货架上所有架板的精准重量值的实时变化情况推断出有哪一个货品被取走或被放回，不但能判断被取走或被放回货品的种类，还能准确判断该货品的身份并获取其货品编号。
即使在一个货架上的同一个托盘内放置多种不同种类的多件货品，一旦托盘重量发生变动，计算机也可准确识别出被取走或放回的货品对应的货品编号，进而准确地更新用户的购物信息。即使出现用户错拿错放货品的情况，该用户的购物记录也不会出错，即使有货品被人摆放至错误位置，也不会影响下一个用户的购买过程，从根本上解决了错拿错放问题，提升了用户体验。
附图说明
图1为本发明实施例所述的无人超市俯视图；
图2为本发明实施例所述货架的整体结构示意图；
图3为本发明实施例所述的托盘及架板的结构示意图；
图4为本发明实施例所述货品识别方法的流程图；
图5为本发明实施例所述货品数据库构建步骤的流程图；
图6为本发明实施例所述托盘变化值第一采集步骤的流程图；
图7为本发明实施例所述托盘变化值第二采集步骤的流程图；
图8为本发明实施例所述货品识别系统的结构示意图。
具体实施方式
以下参考说明书附图完整介绍本发明的优选实施例，使其技术内容更加清楚和便于理解。本发明可以通过许多不同形式的实施例来得以体现，其保护范围并非仅限于文中提到的实施例。
在附图中，结构相同的部件以相同数字标号表示，各处结构或功能相似的部件以相似数字标号表示。本发明所提到的方向用语，例如，上、下、前、后、左、右、内、外、上表面、下表面、侧面、顶部、底部、前端、后端、末端等，仅是附图中的方向，只是用来解释和说明本发明，而不是用来限定本发明的保护范围。
当某些部件被描述为“在”另一部件“上”时，所述部件可以直接置于所述另一部件上；也可以存在一中间部件，所述部件置于所述中间部件上，且所述中间部件置于另一部件上。当一个部件被描述为“安装至”或“连接至”另一部件时，二者可以理解为直接“安装”或“连接”，或者一个部件通过一中间部件间接“安装至”或“连接至”另一个部件。
本实施例涉及一种货品识别系统，是用于无人超市或无人便利店的无人售货系统中的一部分。
如图1～图3所示，所述无人售货系统包括一封闭空间1，其内设有多个货架2，每一货架2包括支架3及可拆卸式安装在支架3上的多个架板4，多个架板4在不同高度彼此平行或者在同一高度彼此平齐。每一架板4上设有一个托盘5或多个并列设置的托盘5，每一托盘5上放置有至少一种货品，每一种货品的数目为一个或多个。本实施例中，在同一托盘5上的货品种类相同或不同，优选同一托盘5上仅摆放一种货品。
如图2、图3、图9所示，本发明所述货品识别系统100，包括电子设备10、至少一货 架2以及至少一微重量传感器6。
每一托盘5与架板4之间都设置有至少一个微重量传感器6，优选近似于长方体形状的重量传感器，其一端的下表面连接至一架板4，其另一端的上表面连接至一托盘5，每一个微重量传感器6连接至电子设备10，实时获取其上方的托盘5及被放置在托盘5上表面的全部货品总的精准重量值并将该精准重量值传送至电子设备10。
电子设备10优选一计算机，其存储器中存储有一货品数据库，在所述货品数据库中存储有至少一货品的货品信息；所述货品信息包括每一种类货品的货品种类、货品价格及标准重量值。进一步地，所述货品信息还包括每一个货品的货品编号，根据货品编号可以查找每一货品的信息。在该货品数据库中还存储有每一托盘的编号m，每一托盘在空置状态下的精准重量值Pm0及摆放在每一托盘上的货品种类，每一个货品的精准重量值，每一个货品的摆放位置等。同一种类的货品具有相同的标准重量值，但由于个体差异，同一种类的不同货品具有不同的精准重量值。
所述精准重量值精确到10 x克，x为整数，-3≤x≤0，优选x＝-1或-2，显然精度更高的传感器成本更高，出于成本考虑，传感器精度只要可以判断出货品个体即可，便利店的经营者可以根据货品总数选择合适的传感器精度。当有某一货品被用户取走时，该货品所处托盘5下方的微重量传感器6采集的精准重量值数据就会变小，当有某一货品被放回时，该货品所处托盘5下方的微重量传感器6采集的精准重量值数据就会变大。电子设备10通过微重量传感器6可以实时监控每一托盘的精准重量变化值，判断是否有货品被取走或被放回，判断该货品的货品编号，记录该货品的货品信息。
在理想状态下，用户素质都比较高，每一次放回货品时都能将该货品正确放回至该货品原属货架。或者，在无人超市墙壁上提醒用户，错放货品会导致购物记录出错，支付金额会超过实际消费金额，迫使所有用户在每一次放回货品时都能将该货品正确放回至该货品原来所处的位置。但此种方法始终未能从根本上解决了用户在购物中因错拿错放货品导致其购物记录出错的问题，如果用户将货品放错位置，导致系统记录出错，就会给客户较差的用户 体验。
本实施例中，同一种类的货品具有相同的标准重量值，但由于个体差异，通常一件货品的精准重量值与其标准重量值会有一定的偏差，该偏差一般是小于或等于5％。事实上，世界上没有绝对相同的两件货品，重量精度越高，一件货品与其他货品重量相同的可能性也就越小。
假设一个便利店有多件货品的精准重量值在精确到1克时完全相同，当微重量传感器的重量精度为0.1克时，出现两件货品精准重量值完全相同的可能性是10％；当微重量传感器的重量精度为0.01克时，出现两件货品精准重量值完全相同的可能性是1％。
事实上，即使微重量传感器的重量精度只精确到1克，相同种类的几个不同货品也很可能具备不同的重量值，出现两件以上货品的精准重量值完全相同的可能性会进一步降低。举例来说，若一件货品的标准重量值为100克，其重量值合理偏差范围是正负5％，该种类下的每一件货品的真实重量值可能为95～105克之间的任意一个数值，出现两件货品精准重量值完全相同的可能性是1/11，即9.1％。在这种情况下，当微重量传感器的重量精度精确到0.1克时，出现两件货品精准重量值完全相同的可能性是0.91％，当微重量传感器的重量精度精确到0.01克时，出现两件货品精准重量值完全相同的可能性是0.091％。
一般来说，一个普通便利店被摆放在货架上的货品总数量为300-3000件，被摆放在同一托盘上的同种类的货品数目一般只有5-30个，因此便利店出现两件商品的精准重量完全相同的可能性很小，可以忽略不计。
现有技术的微重量传感器的重量精度有些已经可以做到10 -4克，重量传感器的精度越高，出现两件商品的精准重量值完全相同的可能性越小。通常情况下，计算机可以忽略小概率事件的影响，认为每一件货品都具有唯一的精准重量值。因此，在本实施例中，货品的精准重量值可以被用作为每一件货品专有的参数，精准重量值也可以被称为微重量值、微重量参数或微重量属性，在实际应用中，可以根据货品的精准重量值的差异来辨别每一件货品的货品编号。
本发明所述电子设备将每一货品的精准重量值作为货品取放状态的判断依据，根据货架上每一托盘的重量变化情况，准确地判断被取走或被放回货品的货品编号，进而从货品数据库中调用该货品的价格和种类等货品信息。在判断出取放货品的用户身份后，实时更新该用户的购物数据库。
一般情况下，为方便用户查找货品，每一托盘只放一种货品；当然，也可以放置种类相似的货品，如几种不同品牌、不同规格、不同价格的饮料可以被放在同一托盘上，并不会影响用户体验。
当某一托盘上的某一货品被取走时，该货品所处托盘下方的微重量传感器实时采集该托盘的精准重量值，与货品数据库中该托盘前一时刻的精准重量值对比并计算二者的差值，获取该托盘的精准重量变化值。由于在货品数据库中预先存储有每一托盘的编号及摆放在每一托盘上的货品的种类，电子设备10根据重量发生变化的托盘的编号即可获取被取走货品的可能的种类。由于在货品数据库中预先存储有每一种类货品的标准重量值和每一件货品的精准重量值，电子设备10根据该托盘的精准重量变化值即可判断被取走货品的货品种类和货品编号。由于每一个货品对应一个微重量参数，三千个货品会有三千个精准重量值，即使同一个托盘内被放置有多种不同种类的货品，所述电子设备也可准确识别出被拿取的货品对应的货品编号。在该货品的编号被识别的同时，电子设备10将该托盘的新的精准重量值存储至所述货品数据库，以替换该托盘的原精准重量值，完成货品数据库更新。同时，电子设备10将被取走货品的货品信息写入取走该货品的用户的购物数据库，以实时更新该用户的购物记录。
当用户将一货品被放回原位置时，电子设备也可准确识别出被放回的货品对应的货品编号，其数据处理流程与前一过程类似。电子设备将该托盘的新的精准重量值存储至所述货品数据库，以替换该托盘的原精准重量值，完成货品数据库更新。同时，电子设备10将被放回货品的货品信息从放回该货品的用户的购物数据库中删除，以实时更新该用户的购物记录。
如果发生错拿错放现象，有用户将原属于A托盘的货品放置于B托盘，由于A、B托盘都是连接至同一台电子设备10中，因此该计算机还是能根据B托盘精准重量变化值识别出该货品的精准重量值，并据此判断其货品编号。在确定放回该货品的用户身份和该货品的编号后，从该用户的购物数据库中删除这一货品对应的货品信息。如图4所示，本发明提供一种货品识别方法，包括如下步骤S1-S12。
S1、货架设置步骤，设置至少一货架，每一货架包括多个彼此独立的托盘。
S2、传感器设置步骤，设置至少一微重量传感器于每一托盘下方，用以实时监控托盘重量的变化。微重量传感器测得的精准重量值精确到10 x克，x为整数，-3≤x≤0。
S3、货品摆放步骤，摆放至少一货品于至少一托盘内，每次仅摆放一个货品。同一托盘上不同货品的种类可以相同，也可以不同。在摆放货品时，工作人员依次摆放每一货品，每次摆放一件，以便电子设备采集每一货品的精准重量值，即微重量参数。
本实施例中，工作人员在摆放货品前，在电子设备(计算机)中设置每一托盘上应当摆放的货品种类，同时录入每一种货品的若干项货品信息，如货品标准重量值、货品种类、货品产地及货品价格等。
如果某些托盘没有被设置货品种类，可以通过扫码方式补充设置。例如，在每一货品的外包装上设有条形码或二维码，记录有该货品的若干项货品信息。当工作人员需要将货品摆放至一个没有被设置货品种类的托盘上时，在摆放货品前，先用扫码装置(如条形码读写器或二维码读写器)扫描一货品的条形码或二维码，获取并传送该货品的货品信息至所述电子设备(计算机)货品数据库。扫码后，工作人员依次摆放同一种类的货品至某一托盘，电子设备获取发生重量变化的托盘的编号，设置该托盘上应当摆放的货品种类即为前一时刻扫码记录下来的货品种类。
S4、货品数据库构建步骤，构建货品数据库，在所述货品数据库中存储有至少一货品的货品信息；所述货品信息包括每一种类货品的货品种类、货品价格及标准重量值。进一步地，所述货品信息还包括货品编号，与每一个货品一一对应，便于计算机查找货品。可选择 地，在该货品数据库中还记录有每一托盘的托盘编号m，每一托盘在空置状态下的精准重量值Pm0，每一个货品的精准重量值，每一个货品的摆放位置等。所述精准重量值精确到10 x克，x为整数，-3≤x≤0；同一种类的货品具有相同的标准重量值，但由于个体差异，同一种类的不同货品具有不同的精准重量值。
货品数据库构建步骤S4与货品摆放步骤S3同步进行，在货品摆放的同时采集一次该货品的精准重量值；货品摆放过程中，每摆放一件货品，货架的微重量传感器获取该托盘的实时重量变化值，传送至所述电子设备(计算机)，电子设备会实时获取一个重量感应变化值，并将其记录为该货品的精准重量值，存储在货品数据库中，计算机根据货架发生重量变动的次数可以记录每一货品的编号，每变动一次，计数器加一，根据发生重量变动的托盘编号可以获取每一货品被摆放的托盘位置。
即使在一个货架上的同一个托盘内放置多种不同种类的货品，计算机也可准确识别出被拿取或被放回的货品的货品编号，从而不会出现因错拿错放货品导致其购物记录出错的问题，并且即使上一位用户出现错拿错放货品的现象，也不会影响到下一位用户的正常购物结算，从根本上解决了用户在购物中因错拿错放货品导致其购物记录出错的问题。
具体的，如图5所示，所述货品数据库构建步骤S4，具体包括如下步骤S41-S46。S41、货品数据库初始化步骤，初始化货品数据库，录入货架编号、托盘编号及每一托盘的精准重量值；同时，可以在货品数据库中预先设置每一货架的每一托盘上摆放货品的种类。在初始状态下，所述货品数据库中每一托盘的精准重量值为该托盘在无货品状态下的精准重量感应值。S42、货品种类录入步骤，每一货品外部二维码或条形码上存储有货品信息，每个货品被摆放在货架前都要先被扫一次码，通过扫码的方式录入每一货品的货品信息至所述货品数据库，所述货品信息包括该货品的货品种类、该种类货品的标准重量值以及货品价格，也可以包括货品编号。S43、托盘变化值第一采集步骤，一货品被摆放至一货架的一托盘上之后，实时采集每一托盘的实时精准重量变化值Qmi。m为托盘编号，i为货品摆放过程中每一所述托盘的重量变化次数，可以根据重量变化次数i计算被摆放货品的编号。S44、托盘变化 值第一判断步骤，判断所有托盘的精准重量变化值Qmi是否全部为0；若是，返回货品种类录入步骤S42；若否，执行下一步骤。S45、货品重量记录步骤，当一托盘的精准重量变化值Qmi为非零数值时，计算该精准重量变化值Qmi与该种类货品的标准重量值A的差值(Qmi-A)；当该差值的绝对值|Qmi-A|与该种类货品的标准重量值A的比值小于一预设阈值(优选3％-5％)时，在所述货品数据库中记录该托盘的精准重量变化值Qmi为该货品的精准重量值；S42中所述货品信息还包括每一货品的精准重量值Qmi。所述预设阈值为该种类货品的标准重量值A的某一偏差比例，如A的3％-5％，其为人工设置方式，可根据实际情况变更数值，防止非该种类货品录入错误，防止该种类货品出现严重的质量问题。S46、货品数据库第一更新步骤，录入该托盘的实时精准重量感应值至所述货品数据库，以替换所述货品数据库中该托盘的精准重量值；返回所述货品种类录入步骤S42。
在货品数据库构建步骤S4中，能够准确的将每一货品的货品信息录入至所述货品数据库中，尤其是能够实现每一货品的精准重量值Qmi对应一个货品编号，一个货品编号对应一个货品价格，通过根据货架的托盘的精准重量变化值与所述货品数据库中货品的精准重量值Qmi对比，便可识别出被拿取或被放回的货品对应的货品编号，这种方式完全依靠每一货品的精准重量值(微重量参数)来识别货品，与货品的摆放位置无关，从而不会出现因错拿错放货品导致其购物记录出错及购物结算出错的问题，从根本上解决了用户在购物中因错拿错放货品导致其购物记录出错的问题。
在步骤S46中，货品数据库中托盘的精准重量值数据被更新之后，返回所述货品种类录入步骤S42，每摆放一件货品，重复执行一次步骤S42-S46。这种录入方式可以准确的对应每一货品编号录入一个货品的精准重量值。
在另一改进实施例中，工作人员在摆放每一种类货品前，先用扫码装置扫描一货品的条形码或二维码，获取并传送该种类货品的货品信息至所述电子设备(计算机)，存储在货品数据库中。之后，工作人员依次摆放该种类的多个货品，货架实时获取每一货品的精准重量值，存储在货品数据库中。
步骤S42中，录入一种货品的货品信息至所述货品数据库，所述货品信息包括货品种类、该种类货品的标准重量值以及货品价格，但不包括该货品的货品编号。在步骤S43中，该种类的多个货品被依次摆放至一货架的一托盘上，由货架根据托盘重量值每一次变动情况记录每一货品的货品编号及精准重量值，同种类货品的其他货品信息皆相同。这种录入方式可以通过货品种类进行依次录入同一种类货品的每一精准重量值，提高了录入货品的精准重量值的效率，并且同一种类货品的货品价格相同，不会出现结算出错的问题，同时通过由货架根据托盘重量值变动情况记录每一货品的货品编号，可省去人工登记货品编号的步骤，进一步提高了录入货品信息的效率。
如图6所示，本实施例中，所述托盘变化值第一采集步骤S43，具体包括如下步骤S431-S432。S431、托盘重量第一采集步骤，在一货品被摆放至一货架的一托盘上之后，采集每一托盘的实时精准重量感应值Pmi，m为托盘编号，i为货品摆放过程中每一所述托盘的重量变动次数。S432、托盘重量差第一计算步骤，计算该货架上的每一托盘的精准重量变化值Qmi，其为每一托盘的实时精准重量感应值与所述货品数据库中该托盘的精准重量值的差值，即Qmi＝Pmi-Pm(i-1)。由于每一托盘在空置状态下的精准重量值Pm0(m为托盘编号)是货品数据库中已知的，根据递归法，可以计算出每一货品的精准重量值。
由于每一种货品的价格相同，因此在执行步骤S431-S432时，可首先获取新增货品的种类，再基于新增货品的种类连续获取多个新增货品的每一精准重量值的差值Qmi作为每一新增货品的精准重量值录入所述货品数据库，即可快速录入新增货品的精准重量值信息，也不会影响货品价格的结算。
在另一实施例中，所述货品数据库构建步骤S4还可以为人工录入方式，工作人员利用高精度电子秤，依次测量并记录每一货品的精准重量值；同时可以人工录入货品信息，如货品编号、货品种类及货品价格等；还可以录入摆放每一货品的货架编号及托盘编号。
托盘变化值第一采集步骤S43、托盘变化值第二采集步骤S5，实时采集一货架上的每一托盘的精准重量变化值。
如图7所示，本实施例中，所述托盘变化值第二采集步骤S5，具体包括如下步骤S51-S52。S51、托盘重量第二采集步骤，实时采集一货架上的每一托盘的实时精准重量感应值。S52、托盘重量差第二计算步骤，实时计算每一托盘的精准重量变化值Qni，其为每一托盘的实时精准重量感应值与所述货品数据库中该托盘的精准重量值的差值，Qni＝Pni-Pn(i-1)。其中，实时精准重量感应值Pni，n为托盘编号，i为货品摆放过程中每一所述托盘的重量变动次数。这样可以依次获取每一托盘的精准重量变化值Qni，通过将与通过根据货架的托盘的精准重量变化值Qni与所述货品数据库中货品的精准重量值Qmi对比，便可识别出被拿取或被放回的货品的货品编号，这种方式完全依靠每一货品的精准重量值(微重量参数)来识别货品，与货品的摆放位置无关，从而不会出现因错拿错放货品导致其购物记录出错及购物结算出错的问题，从根本上解决了用户在购物中因错拿错放货品导致其购物记录出错的问题。若托盘及货品的重力在极短时间内发生快速变化时，重量传感器的感应数据也会快速发生变化，无法采集到准确的重量值，只有在稳定状态下，重量传感器才能采集到准确的重量感应值。在步骤S51中，当一托盘的重量感应值发生变化时，判断重量传感器是否能采集到准确的、稳定的感应数据，若是，执行步骤S52；若否，判断该托盘为异常状态，也可以称之为不稳定异常状态；其原因有可能是用户身体或随身物品倚靠在货架上，有可能是用户将货品在托盘上拖动或按压，也有可能是有昆虫在托盘上爬动，等等。这些情况可以认为是托盘状态发生异常，发出报警信号，等待一定时间(一预设时间段，如1～3秒)后，若异常状态消除，重新执行步骤S51。其中，报警信号可以是蜂鸣器的声音、显示屏上异常提醒或异常信息记录，用以提醒管理人员或用户。
S6、托盘变化值第二判断步骤，判断所有托盘的精准重量变化值Qni是否全部为0；若是，返回所述托盘变化值第二采集步骤S5；若否，执行下一步骤。一托盘的精准重量变化值即为该托盘上被取走或被放回的货品的精准重量值，若所有托盘的精准重量变化值全部为0，说明没有任何货品被取走或被放回，返回步骤S5，准备下一次采集；若某一托盘的精准重量变化值为非零数值，可以判断是发生货品或异物被取走或被放回的情况，其中异物是 指不属于该便利店原有货品的物品。S7、取放状态判断步骤，判断该托盘的精准重量变化值Qni是否小于0，若是，判断有货品从该托盘被取走，若否，判断有货品或物品被放回至该托盘。
S8、货品判断步骤，当一托盘的精准重量变化值Qni为非零数值时，根据该精准重量变化值识别是否有货品被取走或被放回，获取该货品对应的货品编号。判断该托盘的精准重量变化值是否小于0；若是，判定有货品被从该托盘取走；若否，判定有物品或货品被放回至该托盘。
在步骤S8中，判断该精准重量变化值的绝对值|Qni|是否与所述货品数据库中一货品的精准重量值相同；若是，判断该货品即为被取走或被放回的货品，获取其货品编号。若否，说明有用户将不属于便利店的某一物品或者放置在托盘上，可以判断该托盘为异常状态，可以称之为稳定异常状态；可选择地发出前述的报警信号，提醒用户或便利店管理者消除异常，返回所述托盘变化值第二采集步骤S5。
本实施例中，所述货品数据库包括每一货品所处托盘的托盘编号；同一托盘上被放置同一种类的货品。所述货品判断步骤S8包括如下两种情况：当一托盘的精准重量值为负值时，根据其托盘编号从所述货品数据库中调用该托盘上每一货品的精准重量值；判断该精准重量变化值的绝对值与该托盘上一货品的精准重量值是否相等，若是，判定该货品为被取走的货品，获取该货品的货品编号。当一托盘的精准重量值大于零时，判断放回该货品的用户的身份；调用放回该货品的用户的购物数据库；判断该精准重量变化值的绝对值是否与该用户的购物数据库中一货品的精准重量值相同；若是，判定该货品为被放回的货品，获取该货品的货品编号。
S9、货品数据库第二更新步骤，存储该托盘的实时精准重量感应值至所述货品数据库，以替换所述货品数据库中该托盘的精准重量值，返回所述托盘变化值第二采集步骤S5，准备监控每一货架下一次的货品变动情况。如果一托盘的重量感应值发生不稳定变化，货品数据库中该托盘的实时精准重量感应值无需更新；如果一托盘的重量感应值发生稳定变化，无 论其变化值是否对应货品数据库中的某一货品，无论被取走或被放回的物品是否为该便利店的原有货品，货品数据库中存储的相应托盘的精准重量感应值都需要被实时更新。
本实施例的异常状态是指托盘的实时重量感应值发生意外变化的情况，包括不稳定异常状态和稳定异常状态。不稳定异常状态下，重量传感器无法获取稳定的实时重量感应值，其形成原因可能是有人刻意在某货品上施加向上的力、有人将货品在托盘上拖动、有人倚靠或按压托盘、有昆虫在托盘上停留或爬动等。不稳定异常状态维持时间比较短，因此，可以在等待一段时间后，等异常状态消除后，重新采集重量感应值。稳定异常状态下，重量传感器可以获取稳定的实时重量感应值，但此时托盘重量变化的差值与货品数据库中的任一商品都不能形成匹配，电子设备无法确认该货品的身份，并将该货品从用户的购物数据库删除。稳定异常状态的形成原因可能是用户将不属于便利店的某一物品放置在托盘上，或者将吃过一部分的食品重新放回托盘。稳定异常状态维持时间比较长，这种情况下，托盘的实时重量感应值可以被存储至货品数据库，以替换所述货品数据库中该托盘的精准重量值。无论异常状态是否消除，当该托盘的其他货品被用户取放时，电子设备依然可以记录正确的购物行为，不会影响到其他用户的体验。
如图4所示，本实施例中，所述货品识别方法还包括如下步骤S21-S24。
S21、用户身份识别步骤，识别每一用户的身份信息。在一封闭空间1内设置前文所述的货架及托盘，将封闭空间打造成无人便利店。在封闭空间1的出入口处设置闸机，闸机上设置身份识别装置，本实施例优选二维码读写装置，读取用户手机APP中的身份二维码，识别出用户身份。用户可以在手机APP中绑定电子支付账号，并授权给便利店经营者，允许经营者扣款。
S22、购物数据库生成步骤，根据所述用户的身份信息生成该用户的购物数据库。所述购物数据库为对应每一用户的购物信息，包括用户的身份信息以及被该用户取走的货品对应的货品信息，包括被取走货品的种类、数量及价格，以便于便利店经营者计算用户需要支付的金额，以便为客户结算。
S23、购物用户判断步骤，当有货品被取走或被放回时，判断取走或放回该货品的用户的身份。无人便利店的封闭空间内设有大量三维摄像头，从用户在入口扫码时开始实时监控每一用户在便利店中的位置，获取每一用户的实时三维坐标集。当一托盘重量发生变动且同时用户的手部与货架发生互动时，判定是该用户取走或放回了货品。用户的手部与货架发生互动是指，计算机判断出有一用户手部坐标与重量变动的托盘坐标重合，或者，判断出有一用户手部在重量变动的托盘附近出现，等等。
S24、购物数据库更新步骤，在每次货品判断步骤S8之后，实时更新相应用户的购物数据库。当有货品被取走时，根据所述货品编号从所述货品数据库中调用该货品的货品信息，将该货品的货品信息记录在取走该货品的用户购物数据库中；所述货品信息包括货品种类、该种类货品的标准重量值以及货品价格等；当有货品被放回至货架时，从放回该货品的用户的购物数据库中删除该货品的货品信息。购物数据库更新步骤根据每一用户的实时购物行为，实时更新每一用户的购物记录，使得每一用户的实际购物情况与其电子购物车中的购物记录完全一致，以便用户结算。
如果便利店规模比较大，货品较多，整个便利店中有可能出现精准重量值完全相同的两个货品的概率也会比较大，计算机识别出错率就会被提升。现有的重量传感器虽然可以采集到高精度的重量值，但是传感器的精度越高，其制造要求越高，其成本越高、价格越贵，而本发明中需要大量的重量传感器，会给无人便利店的经营者带来较大的成本压力。
上述实施例的理论依据在于，一家便利店的每一货品都具有唯一的精准重量值(微重量参数)，计算机可以据此有效区分每一件货品。然而，一家便利店的数百件或数千件货品中，出现两个以上不同种类货品的精准重量值相同的情况，虽然是小概率事件，但理论上还是有可能发生，因此，上述实施例对重量传感器的精度要求较高，精准重量值精确到10 -2克或者10 -3克，以确保每一货品都具有唯一的精准重量值。由于整体解决方案需要使用大量的重量传感器，高精度传感器的硬件成本较高，会造成整个解决方案的成本过高，影响本专利的推广应用。
在另一改进实施例中，电子设备10可以先确定放回该货品的用户的身份，可以采用门禁处扫码识别用户身份的方案及三维相机实时监控用户位置的方案。当已知身份的某一用户的手部坐标集与一托盘上方坐标集重合，且同时该托盘的重量感应值发生变动，则可以判断取走或放回该货品的用户的身份。
在所述改进实施例中，在S8货品判断步骤中，当一托盘的精准重量变化值小于零时，电子设备(计算机)可以判断出有货品被取走；由于之前摆放在每一托盘上的货品种类是预存在货品数据库中的，电子设备(计算机)可以判断出该货品的种类。例如，一托盘上摆放同一种类、同一品牌、重量近似相同的货品；当该托盘重量变轻了，计算机就可以确定被取走的货品的种类；进一步地，根据精准重量变化值与货品数据库中预存的该托盘上每一货品的精准重量值的对比，判断并记录该货品的货品编号。进一步地，在购物用户判断步骤S23中判断出取走该货品的用户的身份，在购物数据库更新步骤S24中将该货品的货品信息记录在至取走该货品的用户的购物数据库中。
在所述改进实施例中，在S8货品判断步骤中，当一托盘的精准重量变化值大于零时，S8货品判断步骤依次包括取放状态判断步骤，购物用户判断步骤，购物数据库调用步骤以及物品判断步骤。电子设备(计算机)可以判断出有物品被放回，进而判断出放回该物品的用户的身份，调用放回该物品的用户的购物数据库，判断该精准重量变化值的绝对值是否与该用户的购物数据库中一货品的精准重量值相同；若是，判定该物品为被放回的货品，获取该货品的货品编号；若否，判定该物品并非是该用户之前从货架上取走的货品，并非本店货品，有可能是该用户的随身物品，如手套、手机等。无论被放置在货架托盘上的物品是否为本店货品，在步骤S8后都要执行货品数据库第二更新步骤S9，存储该托盘的实时精准重量感应值至所述货品数据库，以替换所述货品数据库中该托盘的精准重量值，返回所述托盘变化值第二采集步骤S5，准备监控每一货架下一次的货品变动情况。在确定放回该货品的用户身份和该货品的编号后，执行S24购物数据库更新步骤，从该用户的购物数据库中删除这一货品对应的货品信息。
在上述改进实施例中，一个用户在便利店的采购量一般仅为几个或十几个，最多不过几十个，这几十个货品分属于不同种类，其中刚好包括微重量值完全相同的两件货品的概率会小于千万分之一，完全可以忽略不计，因此每一货品的精准重量值只需要精确到1克或0.1克的精度即可。在这一方案中，可以利用精度较低、成本较低的重量传感器，有效节省运营和维护成本，进一步降低计算机出错的概率，提升识别精度。
用户在无人便利店内，其购物数据库可能被一次或多次更新，在至少一次的购物数据库更新步骤S24之后，还包括一结算步骤，根据一用户的购物数据库计算该用户需要支付的金额。当一用户完成购物，离开便利店时，该便利店内的多个摄像头拍摄不到该用户，电子设备(计算机)会认为该用户已经离场，根据该用户的购物数据库计算该用户的消费金额，由于用户之前在手机APP中绑定过电子支付账号并授权允许便利店经营者扣款，便利店经营者可以直接从该用户的电子支付账户中扣除相应款项。用户无需排队结算，完成购物可以直接离场，获得非常好的用户体验。
如图9所示，本实施例所述的电子设备10，包括至少一存储器11以及至少一处理器12。存储器11内设有一种存储介质，用于存储可执行程序代码。处理器12连接至存储器11，通过读取所述可执行程序代码，来运行与所述可执行程序代码对应的计算机程序，以执行所述货品识别方法中的步骤S4-S9，步骤S21-S24中的至少一个步骤。
本发明的有益效果在于，提供一种货品识别系统及货品识别方法、存储介质及电子设备，可以通过实时监测货架上的每一托盘的精准重量变化值，由货架上所有架板的精准重量变化推断出有哪一种货品被取走或被放回，可以判断被取走或被放回货品的货品编号，即使在一个货架上的同一个托盘内放置多种不同种类的货品，也可准确识别出被拿取或放回的货品的编号，从而对于用户的购物信息判断更加精准，不会出现因错拿错放货品导致其购物记录出错的问题，提升了用户体验，并且即使上一位用户出现错拿错放货品的现象，也不会影响到下一位用户的正常购物结算，从根本上解决了用户在购物中因错拿错放货品导致其购物记录出错的问题。

Claims (14)

  1. 一种货品识别方法，其特征在于，包括如下步骤：
    构建货品数据库，所述货品数据库包括至少一货品的精准重量值及至少一托盘的精准重量值；所述精准重量值精确到10 x克，x为整数，-3≤x≤0；实时采集一货架上的每一托盘的精准重量变化值；
    当一托盘的精准重量变化值为非零数值时，根据该精准重量变化值识别是否有货品被取走或被放回，获取该货品对应的货品编号；
    以及
    存储该托盘的实时精准重量感应值至所述货品数据库，以替换所述货品数据库中该托盘的精准重量值。
  2. 如权利要求1所述的货品识别方法，其特征在于，还包括如下步骤：
    识别每一用户的身份信息；
    根据所述用户的身份信息生成该用户的购物数据库；以及
    当有货品被取走或被放回时，判断取走或放回该货品的用户的身份。
  3. 如权利要求1所述的货品识别方法，其特征在于，
    在获取该货品对应的货品编号的步骤之后，还包括如下步骤：
    根据所述货品编号从所述货品数据库中调用该货品的货品信息；以及
    当有货品从一托盘被取走时，将该货品的货品信息记录在取走该货品的用户的购物数据库中；当有货品被放回至一托盘时，从放回该货品的用户的购物数据库中删除该货品的货品信息。
  4. 如权利要求1所述的货品识别方法，其特征在于，
    在根据该精准重量变化值识别是否有货品被取走或被放回的步骤中，
    判断该托盘的精准重量变化值是否小于0；若是，判定有货品被从该托盘取走；若否，判定有物品或货品被放回至该托盘。
  5. 如权利要求1所述的货品识别方法，其特征在于，
    在获取该货品对应的货品编号的步骤中，
    判断该托盘的精准重量变化值的绝对值是否与所述货品数据库中一货品的精准重量值相等；若是，判定所述货品数据库中的该货品即为被取走或被放回的货品，获取其货品编号。
  6. 如权利要求1所述的货品识别方法，其特征在于，
    获取该货品对应的货品编号的步骤，包括如下步骤：
    当一托盘的精准重量值大于零时，判断放回该物品的用户的身份；
    调用放回该物品的用户的购物数据库；
    判断该精准重量变化值的绝对值是否与该用户的购物数据库中一货品的精准重量值相同；若是，判定该物品为被放回的货品，获取该货品的货品编号。
  7. 如权利要求1所述的货品识别方法，其特征在于，
    构建货品数据库的步骤，具体包括如下步骤：
    初始化货品数据库，录入货架编号、托盘编号及每一托盘的精准重量值；
    在初始状态下，所述货品数据库中每一托盘的精准重量值为该托盘在无货品状态下的精准重量感应值；
    录入一货品的货品信息至所述货品数据库，所述货品信息包括该货品的货品编号、货品种类、该种类货品的标准重量值以及货品价格；
    一货品被摆放至一货架的一托盘上之后，实时采集每一托盘的精准重量变化值；
    当一托盘的精准重量变化值为非零数值时，在所述货品数据库中记录该托盘的精准重量变化值为该货品的精准重量值；以及
    录入该托盘的实时精准重量感应值至所述货品数据库，以替换所述货品数据库中该托盘的精准重量值。
  8. 如权利要求7所述的货品识别方法，其特征在于，
    在所述货品数据库中记录该托盘的精准重量变化值为该货品的精准重量值的步骤中，
    当一托盘的精准重量变化值为非零数值时，计算该精准重量变化值与该种类货品的标准重量值的差值；
    当该差值的绝对值与该种类货品的标准重量值的比值小于一预设阈值时，在所述货品数据库中记录该托盘的精准重量变化值为该货品的精准重量值。
  9. 如权利要求1所述的货品识别方法，其特征在于，
    实时采集一货架上的每一托盘的精准重量变化值的步骤，具体包括如下步骤：
    采集每一托盘的实时精准重量感应值；以及
    计算该货架上的每一托盘的精准重量变化值，其为每一托盘的实时精准重量感应值与所述货品数据库中该托盘的精准重量值的差值。
  10. 如权利要求9所述的货品识别方法，其特征在于，
    在采集每一托盘的实时精准重量感应值的步骤中，
    当一托盘的重量感应值发生变化时，判断重量传感器是否能采集到准确的、稳定的感应数据，若否，判断该托盘为异常状态，可选择地发出报警信号。
  11. 如权利要求1所述的货品识别方法，其特征在于，
    在实时采集一货架上的每一托盘的精准重量变化值的步骤之后，
    判断一托盘精准重量变化值的绝对值是否与所述货品数据库中一货品的精准重量值相同；若否，判断该托盘为异常状态，可选择地发出报警信号。
  12. 如权利要求1所述的货品识别方法，其特征在于，
    在构建货品数据库的步骤之前，还包括如下步骤：
    设置至少一货架，每一货架包括多个彼此独立的托盘；
    设置微重量传感器于每一托盘下方，所述微重量传感器的重量感应值精确 到10 x克，x为整数，-3≤x≤0；以及
    摆放至少一货品于至少一托盘内，每次仅摆放一个货品。
  13. 一种电子设备，包括：
    存储器，用于存储可执行程序代码；以及
    处理器，连接至所述存储器，通过读取所述可执行程序代码，来运行与所述可执行程序代码对应的计算机程序，以执行如权利要求1-11中任一项所述的货品识别方法中的至少一步骤。
  14. 一种货品识别系统，包括：
    如权利要求13所述的电子设备；
    至少一货架，每一货架包括多个彼此独立的托盘；以及
    至少一微重量传感器，设于每一托盘下方，且连接至所述电子设备；
    其中，所述微重量传感器用以获取至少一货品的精准重量值和/或托盘的精准重量值；所述精准重量值精确到10 x克，x为整数，-3≤x≤0。
PCT/CN2021/078316 2020-02-28 2021-02-27 货品识别方法、货品识别系统及电子设备 WO2021170126A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010129658.8A CN111429655A (zh) 2020-02-28 2020-02-28 货品识别方法及货品识别系统、存储介质及电子设备
CN202010129658.8 2020-02-28

Publications (1)

Publication Number Publication Date
WO2021170126A1 true WO2021170126A1 (zh) 2021-09-02

Family

ID=71547356

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/078316 WO2021170126A1 (zh) 2020-02-28 2021-02-27 货品识别方法、货品识别系统及电子设备

Country Status (2)

Country Link
CN (1) CN111429655A (zh)
WO (1) WO2021170126A1 (zh)

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111429655A (zh) * 2020-02-28 2020-07-17 上海追月科技有限公司 货品识别方法及货品识别系统、存储介质及电子设备
CN112529500A (zh) * 2020-12-16 2021-03-19 石家庄邮电职业技术学院(中国邮政集团有限公司培训中心) 一种超市智能货架系统
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
CN113065824A (zh) * 2021-03-31 2021-07-02 北京京东振世信息技术有限公司 物流任务信息的处理方法、系统、装置、介质与电子设备
CN113554807B (zh) * 2021-04-22 2023-06-02 浙江星星冷链集成股份有限公司 配货清单生成装置以及无人售货系统
CN113554809A (zh) * 2021-04-22 2021-10-26 浙江星星冷链集成股份有限公司 一种商品信息更新装置及无人售货系统
CN113554813B (zh) * 2021-04-22 2023-06-06 浙江星星冷链集成股份有限公司 批量上货管理装置以及无人售货系统
CN113554810A (zh) * 2021-04-22 2021-10-26 浙江星星冷链集成股份有限公司 一种无人售货系统
CN113554808B (zh) * 2021-04-22 2023-09-29 浙江星星冷链集成股份有限公司 一种上货管理装置及无人售货系统
CN113421376B (zh) * 2021-08-23 2021-12-28 浙江星星冷链集成股份有限公司 一种无人售货系统

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108332829A (zh) * 2017-12-18 2018-07-27 上海云拿智能科技有限公司 基于重量监测的货品感知系统及货品感知方法
KR101887685B1 (ko) * 2018-01-08 2018-09-10 정진혁 운영 시스템을 포함하는 무인 자판기
CN108765704A (zh) * 2018-05-23 2018-11-06 济南每日优鲜便利购网络科技有限公司 自动售货系统补货控制方法、装置、设备及存储介质
CN108960965A (zh) * 2018-06-06 2018-12-07 虫极科技（北京）有限公司 无人售货货柜、服务器以及购物订单的创建方法
CN110428547A (zh) * 2019-08-09 2019-11-08 四川虹美智能科技有限公司 一种自动售卖机及商品售卖方法
CN111429655A (zh) * 2020-02-28 2020-07-17 上海追月科技有限公司 货品识别方法及货品识别系统、存储介质及电子设备

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108806073B (zh) * 2018-05-31 2020-12-08 深圳正品创想科技有限公司 一种混合货道以及无人售货柜
CN110197416A (zh) * 2019-05-21 2019-09-03 李志高 通过重力识别商品的方法和装置
CN110211285A (zh) * 2019-06-04 2019-09-06 李志高 商品识别方法和装置

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108332829A (zh) * 2017-12-18 2018-07-27 上海云拿智能科技有限公司 基于重量监测的货品感知系统及货品感知方法
KR101887685B1 (ko) * 2018-01-08 2018-09-10 정진혁 운영 시스템을 포함하는 무인 자판기
CN108765704A (zh) * 2018-05-23 2018-11-06 济南每日优鲜便利购网络科技有限公司 自动售货系统补货控制方法、装置、设备及存储介质
CN108960965A (zh) * 2018-06-06 2018-12-07 虫极科技（北京）有限公司 无人售货货柜、服务器以及购物订单的创建方法
CN110428547A (zh) * 2019-08-09 2019-11-08 四川虹美智能科技有限公司 一种自动售卖机及商品售卖方法
CN111429655A (zh) * 2020-02-28 2020-07-17 上海追月科技有限公司 货品识别方法及货品识别系统、存储介质及电子设备

Also Published As

Publication number Publication date
CN111429655A (zh) 2020-07-17

Similar Documents

Publication Publication Date Title
WO2021170126A1 (zh) 货品识别方法、货品识别系统及电子设备
US20200258069A1 (en) Weight monitoring-based sensing system and sensing method
CN108846621B (zh) 一种基于策略模块的库存管理系统
CN109559454B (zh) 基于神经网络识别商品的收银方法以及自助收银台
US20150127496A1 (en) Methods, Devices and Computer Readable Storage Devices for Tracking Inventory
US11908288B2 (en) Imaging used to reconcile cart weight discrepancy
US9836719B2 (en) Consumer demand-based inventory management system
CN108564713A (zh) 一种基于综合测量方式的无人值守零售的商品识别和防作弊系统及其方法
WO2020243984A1 (zh) 货品感知系统、货品感知方法及电子设备
KR20190096975A (ko) 자동 상점 내 등록 시스템
KR20140136089A (ko) 진열 상품 재고 관리 솔루션을 구비한 전자 선반 라벨 시스템
CN111079478B (zh) 一种无人售货架监控方法、装置、电子设备及系统
KR20210047300A (ko) 멀티모달 센서 작동에 기초한 재고의 자동 제품 식별
JP2024040297A (ja) 物品推定装置、物品推定方法、及びプログラム
JP2019109805A (ja) 販売分析装置、販売管理システム、販売分析方法、及びプログラム
WO2021147950A1 (zh) 货品识别方法、货品识别系统及电子设备
JP2021012518A (ja) 無人店舗の商品を管理する管理サーバ及び管理方法
CN111507792A (zh) 一种自助购物方法、计算机可读存储介质及系统
CN212391834U (zh) 一种新型无人售货柜
CN108877039B (zh) 自动售货装置及系统
CN114596661B (zh) 一种多功能智能售货柜
US11900349B2 (en) Self-checkout store
CN111860221A (zh) 一种货柜管理的方法以及货柜
JP2022144490A (ja) 店舗システム及びプログラム
CN114882640A (zh) 一种基于推货售卖机的推货位置识别系统

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: 21760220

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: 21760220

Country of ref document: EP

Kind code of ref document: A1