CN111831673B

CN111831673B - Goods identification system, goods identification method and electronic equipment

Info

Publication number: CN111831673B
Application number: CN202010525545.XA
Authority: CN
Inventors: 马捷昱; 冯立男
Original assignee: Shanghai Chaoyue Moon Technology Co ltd
Current assignee: Shanghai Chaoyue Moon Technology Co ltd
Priority date: 2020-06-10
Filing date: 2020-06-10
Publication date: 2024-04-12
Anticipated expiration: 2040-06-10
Also published as: CN111831673A

Abstract

The invention provides an article identification system, an article identification method and electronic equipment. The goods identification method comprises the following steps: database construction, tray parameter acquisition and weight change value judgment. The invention collects the tray area outline and the tray weight change value of each tray with weight change in real time by installing the pressure blanket on the upper surface of the tray. When the goods are taken off from the tray of the goods shelf or placed on the tray, the type and the quantity of the taken off or placed goods can be judged in real time, so that the shopping records in the shopping database of the user can be ensured to be consistent with the actual shopping situation of the user.

Description

Goods identification system, goods identification method and electronic equipment

Technical Field

The present invention relates to a technology for identifying goods in retail industry, and in particular, to a system, a method and an electronic device for identifying goods.

Background

The Chinese patent with publication number CN108332829A discloses a goods sensing system and a goods sensing method based on weight monitoring, the Chinese patent with publication number CN108497839A discloses a goods shelf capable of sensing goods, and the two patents relate to an intelligent goods shelf for an unmanned convenience store, wherein a weight sensor is required to monitor the change condition of each goods on the goods shelf in real time, and the variety and quantity of goods change are recorded. The scheme is that the key type trays are adopted, a plurality of independent trays are arranged in parallel on each layer, the same kind of goods are placed in each tray, and the plurality of trays on the same layer are not mutually influenced. When the goods are taken out from the goods shelves, the weight sensor can accurately judge the types and the amounts, and the disadvantage is that when the user puts the goods into the wrong position, the computer system can only judge that the goods are put incorrectly, but cannot determine the types and the amounts of the misplaced goods, because various goods have the same weight and are very possible common events, and the weight values of the goods and the types of the goods are not in one-to-one correspondence.

If the user takes down the food on the goods shelf and eat the food, the food after eating is put back to the original tray, the computer system can only recognize that the weight of the food is different from that of the original goods on the tray, and can not judge whether the goods belong to the tray, but the eaten goods or the goods belong to other trays, and can only make simple alarms. If the goods are other goods originally positioned on other trays, the goods can be sold for the second time, and the staff can do not adjust in a short time; if the goods are eaten and cannot be sold for the second time, the alarm should be given to prompt the staff or the user to take away as soon as possible. Meanwhile, the computer can identify which user returns the goods, but because the specific type and quantity of the goods are difficult to judge, the goods cannot be deleted from the electronic shopping cart of the user, so that the shopping record in the electronic shopping cart of the user is inconsistent with the real shopping situation.

In addition, the key type tray structure adopted by the two patents is greatly different from the common shelf tray structure, the mould and the design tray are required to be redeveloped, a large number of weight sensors are installed, and the cost of the key type shelf is much higher than that of the conventional shelf; moreover, the size of each tray of the key type goods shelf is fixed, only some goods with the same size as the tray can be placed, and if other goods with different sizes are placed, the trays must be customized again, so that the use cost of the goods shelf is further increased.

Disclosure of Invention

The invention aims to solve the problems of poor identification accuracy of the types of the put-back goods, wrong taking and misplacement and over-high goods shelf cost of the intelligent goods shelf in the prior art.

In order to achieve the above object, the present invention provides an article identification method comprising the steps of: constructing an article database for storing article information of each article placed on a tray of a goods shelf, wherein the article information comprises the type, the weight of a single article, the column in which each article is positioned and the bottom surface contour; collecting weight change values of each tray in real time; when the weight change value is a non-zero value, recording tray parameters including a weight change value, a column with a changed weight and a tray area contour with a changed weight; judging whether the weight change value is smaller than 0, if so, judging that the goods are taken away from a tray of a goods shelf, judging the type and the quantity of the taken-away goods according to the tray parameters and the goods information, and storing the type, the quantity and the bottom surface contour of the taken-away goods into a shopping database of a user taking the goods; if not, judging that the goods are placed on the tray of the goods shelf, judging the types and the quantity of the goods placed on the goods shelf according to the tray parameters and the shopping database of the user placing the goods, and deleting the goods information corresponding to the goods from the shopping database of the user placing the goods.

Further, the step of judging the type and the number of the removed goods according to the tray parameters and the goods information specifically comprises the following steps: comparing the number of the column with the number of the column in which each article in the article database is positioned, and obtaining the type of the article to be taken; and calculating the quotient of the absolute value of the weight change value and the weight of the single product of the type of the product in the product database, and carrying out rounding processing on the quotient to obtain the number of the removed products.

Further, the step of judging the kind and the number of the goods placed on the shelf comprises the following steps: judging whether the weight change value is the same as the weight of one article or the total weight of a plurality of articles in a shopping database of a user placing the articles; if yes, obtaining M probability results of the types and the numbers of the goods placed on the goods shelves; if not, judging the goods to be foreign matters.

Further, the step of judging the kind and the number of the goods placed on the shelf comprises the following steps: judging whether the contour of the tray area with the weight changed is consistent with the contour of at least one article in a shopping database of a user placing the article; if yes, N possible results of the types and the numbers of the goods are obtained; if not, judging the goods to be foreign matters.

Further, the step of judging the kind and the number of the goods placed on the shelf comprises the following steps: judging whether the weight of the goods is the same as the weight of at least one goods in a shopping database of a user placing the goods; if yes, obtaining M probability results of the types and the quantity of the goods placed on the goods shelf, namely weight comparison results; judging whether the contour of the tray area with the weight changed is consistent with the contour of at least one article in a shopping database of a user placing the article; if yes, N possible results of the types and the quantity of the goods are obtained, namely a contour comparison result; and comparing the weight comparison result with the contour comparison result, wherein if one and only one of the weight comparison result and the contour comparison result are the same, the weight comparison result is the type and the number of the goods placed on the goods shelf.

Further, the article identification method further includes the steps of: identifying identity information of each user; generating a shopping database of each user according to the identity information of each user; when the weight change value is smaller than 0, judging the identity of a user taking away the goods; and when the weight change value is larger than 0, judging the identity of a user placing the goods.

Further, the step of constructing the goods database specifically comprises the following steps: initializing an goods database, and recording the column number of each tray into the goods database; inputting the goods information corresponding to each column to the goods database, wherein the goods information comprises the types and the weights of the goods corresponding to the column; collecting the bottom surface profile of each article, and storing the bottom surface profile into the article database; when an article is placed on a tray, the contour of the area where the weight of the tray changes is the contour of the bottom surface of the article.

Further, the article identification method further includes the steps of: deleting the item information of the taken item from the item database when the weight change value is less than 0; when the weight change value is greater than 0, if the goods placed on the goods shelf are not foreign matters, the goods type, the single-product weight, the column and the bottom surface outline of each goods placed on the goods shelf are stored into the goods database as a group of goods information.

Further, the article identification method further includes the steps of: searching a column corresponding to the goods from the goods database according to the type of the goods placed on the goods shelf; judging whether the column corresponding to the goods of the category is consistent with the column with changed weight; if yes, judging that the goods are placed in the correct columns; if not, judging that the goods are placed in the wrong column, and optionally sending out an alarm signal.

Further, the bottom surface profile of the article includes the shape and size of the bottom surface of the article; and the weight-changing tray region profile includes a shape and size of the weight-changing tray region.

Further, before the step of constructing the item database, the method further comprises the steps of: providing at least one shelf, each shelf comprising a plurality of trays, each tray being separated into at least one column; a pressure blanket is arranged on the upper surface of each tray; placing at least one article in at least one tray, wherein the bottom surface of each article contacts the top surface of the pressure blanket, and each column is placed with the same article; the pressure blanket is used for acquiring weight change values of the tray, columns with weight change and outlines of the tray area with weight change in real time.

The present invention also provides an electronic device including: a memory for storing executable program code; and a processor connected to the memory, for executing a computer program corresponding to the executable program code by reading the executable program code to perform the steps in the article identification method described above.

The invention also provides an article identification system, which comprises the electronic equipment.

Further, the invention also provides an article identification system, which further comprises: at least one shelf, each shelf comprising a plurality of trays, each tray being divided into a plurality of columns for placing goods; at least one pressure blanket attached to the upper surface of each tray and connected to an electronic device; the pressure blanket is used for acquiring weight change values of the tray, columns with weight change and outlines of the tray area with weight change in real time.

Further, the pressure blanket includes a plurality of pressure sensors evenly distributed.

Further, the item identification system also includes a spacer removably mounted over each tray for dividing the tray into a plurality of columns.

The invention has the beneficial effects that the weight and the contour of each type of goods on the tray are acquired and recorded in a mode of installing the pressure blanket on the upper surface of the tray, and the area contour of the tray and the weight change value of the tray, in which the weight of each tray is changed, are acquired in real time.

When the goods are taken away from the tray of the goods shelf, the weight change value of the tray is compared with the goods information corresponding to each tray in the goods database, so that the type and the quantity of the goods taken away from the tray can be judged, and the goods information such as the weight value and the bottom contour of the goods taken away are stored in the shopping database of the user who takes away the goods. When the goods are placed on the tray, the weight change value and the outline of the tray are compared with the weight and the outline of the goods in the shopping database of the user placing the goods, so that the type and the number of the goods placed back to the tray can be judged, and then the goods information corresponding to the goods in the shopping database of the user is deleted, and the shopping record in the shopping database of the user can be ensured to be consistent with the actual shopping situation of the user.

The goods shelf used in the invention has a similar structure to the existing goods shelf, only the pressure blanket is needed to be arranged on the common goods shelf tray, no key type is needed to be specially made, the cost of the goods shelf can be effectively reduced, and the column width of each tray is adjustable and can be used for placing goods with different sizes.

Drawings

FIG. 1 is a schematic view of the overall structure of a pallet according to the present invention;

FIG. 2 is a schematic view of a pressure blanket according to the present invention;

FIG. 3 is a schematic block diagram of the system for identifying goods according to the present invention;

FIG. 4 is a schematic view of the front end structure of the pallet according to the present invention;

FIG. 5 is a schematic view of the rear end structure of the pallet according to the present invention;

FIG. 6 is a flow chart of a method of identifying an item according to the present invention;

FIG. 7 is a flowchart of the item database construction steps of the present invention;

fig. 8 is a flowchart of the user identity acquisition step according to the present invention.

The reference numerals for the various components in the figures are as follows:

1. a bracket, 2, a pressure blanket, 10, a goods shelf, 11, a tray, 12 and a separator,

20. electronic equipment 21, a first carpet face 22, a second carpet face 23, pressure sensors,

30. a camera 41, a first row of holes 42, a second row of holes 51 and a first limit hole,

52. a second limiting hole 100, an article identification system 121, a first bracket 122, a second bracket,

123. third support, 201, memory, 202, processor.

Detailed Description

The following description fully describes the preferred embodiments of the present invention with reference to the accompanying drawings, so that the technical contents thereof are more clearly and easily understood. The present invention may be embodied in many different forms of embodiments and its scope of protection is not limited to the embodiments described herein.

In the drawings, like structural elements are referred to by like reference numerals and like structural or functional elements are referred to by like reference numerals throughout. Directional terms, such as up, down, front, back, left, right, inner, outer, etc., are merely directions in the drawings for illustrating and explaining the present invention, and are not intended to limit the scope of the present invention.

When some element is described as being "on" another element, the element may be directly on the other element; there may also be an intermediate member that is placed on the intermediate member and the intermediate member is placed on another member. When an element is referred to as being "mounted to" or "connected to" another element, it can be directly "mounted to" or "connected to" the other element or be indirectly "mounted to" or "connected to" the other element via an intervening element.

As shown in fig. 1, the present embodiment provides an article identification system 100 for an unmanned convenience store, comprising a shelf 10 placed in the unmanned convenience store, the shelf 10 comprising at least one bracket 1 and at least one pressure blanket 2, each shelf 10 comprising a plurality of trays 11, each tray 11 being divided into a plurality of columns; a pressure blanket 2 is provided on the upper surface of each tray 11, i.e. the pressure blanket 2 is attached to the upper surface of each tray 11.

Specifically, as shown in fig. 2, the pressure blanket 2 includes a first blanket surface 21, a second blanket surface 22, and a plurality of pressure sensors 23; the first blanket face 21 is attached to the upper surface of a tray 11; a second carpet face 22 is arranged above the first carpet face 21, and the edge of the second carpet face 22 is connected to the first carpet face 21; the plurality of pressure sensors 23 are uniformly distributed in the gap between the second blanket surface 22 and the first blanket surface 21. The distance between any two adjacent pressure sensors 23 in the pressure blanket 2 is 2-10 mm, preferably 5mm, so that not only can the real-time change of the weight of the goods on the tray be detected, but also the column position with the changed weight, and the contour of the tray area with the changed weight, namely the contour of the goods taken away or put back, can be detected.

As shown in fig. 3, the article identification system 100 further includes an electronic device 20, the electronic device 20 being provided inside or outside the shelf 10, and the pressure blanket 2 being connected to the electronic device 20. At least one article is placed in at least one tray 11, each column being placed with the same article. The pressure blanket 2 is used to acquire the weight of the goods or the contour of the bottom surface thereof and also to acquire the contour of the pallet area in which the weight is changed. The data of the pressure blanket 2 is transmitted to the electronic device 20, and the electronic device 20 can calculate the reduced weight, and can acquire the column position where the weight is changed and the contour of the goods, so as to judge the type and the number of the goods taken from the goods shelf tray and the type and the number of the goods put back into the goods shelf tray.

To facilitate the separation of the columns, the pallet 10 further comprises a partition 12 removably mounted to the rack 1 for dividing the pallet 11 into at least one column. With the end of the shelf 10 or tray 11 facing the user as the front end and the end facing away from the user as the rear end, it is preferable that the front end of the spacer 12 is lower than the rear end thereof in order to facilitate the picking or placing of goods on the shelf 10 or in the tray 11 by the user. The tray 11 is divided into a plurality of columns for placing different goods, and the shelves 10 may be installed in a refrigerator or freezer.

The embodiment can be directly improved on the basis of the existing tray, the key type keyboard is not needed, the cost can be reduced, and the width of each column is adjustable, so that the keyboard is applicable to scenes with different sizes of goods. According to the embodiment, the specification and the size of the pressure blanket 2 can be customized according to the size of the tray, the traditional goods shelf is transformed into the intelligent goods shelf, and the production cost and the maintenance cost of the goods shelf can be reduced.

As shown in fig. 4 and 5, in the present embodiment, the spacer 12 includes a first bracket 121 and/or a second bracket 122 and/or a third bracket 123; the brackets are all made of at least one metal rod, and the cost is low. The first bracket 121 is detachably mounted to the front end of the tray 11; the second bracket 122 is detachably mounted to the rear end of the tray 11; one end of the third bracket 123 is detachably mounted to the front end of the tray 11, and the other end is detachably mounted to the rear end of the tray 11 to divide the tray 11 into at least one column. Therefore, the columns can be easily separated through the pluggable partition 12, the width of the columns is adjustable, the columns are suitable for the conditions of different sizes of goods, the columns are not strictly limited by the sizes of the goods, the goods shelf 10 does not need to be made into keys, only the pressure blanket 2 is required to be installed on the whole tray 11, and the cost can be reduced.

As shown in fig. 4 and 5, in order to facilitate easy separation of the columns by the pluggable spacer 12, the spacer 12 includes, in addition to the first, second and third brackets 121, 122 and 123, a first row of holes 41 penetrating the front end of the tray 11 and a second row of holes 42 penetrating the rear end of the tray 11. The lower end of the first bracket 121 is inserted into the first row of holes 41, the lower end of the second bracket 122 is inserted into the second row of holes 42, the front end of the third bracket 123 is inserted into the first row of holes 41, the rear end of the third bracket 123 is inserted into the second row of holes 42, the front end and the rear end of the third bracket 123 are respectively installed to the first bracket 121 and the second bracket 122 in a pluggable manner, the first row of holes 41 and the second row of holes 42 are used for fixing the third bracket 123, the third bracket 123 cannot deflect in the front-rear direction, the third bracket 123 is kept relatively stable, the tray 11 can be divided into a plurality of columns by at least one third bracket 123, and each column is used for placing goods of the same kind. The user can pull out at least one third support 123 from the original position at any time and insert the third support 123 into another position, so that the width of each column can be adjusted according to actual requirements, and the width of the column corresponds to the size of the goods to be placed.

As shown in fig. 6, based on the above-described article identification system 100, the present embodiment further provides an article identification method, which includes the following steps S1 to S12. Wherein steps S1-S4 are hardware setting steps.

As shown in fig. 3, the electronic device 20 includes a memory 201 and a processor 202; a storage medium is provided in the memory 201 for storing executable program code; the processor 202 executes a computer program corresponding to the executable program code by reading the executable program code to perform several steps S5-S12 in a method for identifying an item.

In this embodiment, the pressure blanket 2 is installed on the whole tray 11 to obtain the weight change and the contour of the goods, the contour includes the shape and the size, and the contour of the tray area where the weight of each tray 11 is changed is collected in real time, so that the quantity and the kind of the goods can be accurately identified according to the combination of the gravity scheme and/or the visual scheme.

S1, setting at least one shelf 10, wherein each shelf 10 comprises a plurality of trays 11, and each tray 11 is divided into at least one column by a partition 12 detachably connected to the tray 11. The embodiment can be directly improved on the basis of the existing tray, the key type tray is not needed to be made, the cost can be reduced, the field width can be adjusted, and the method is suitable for scenes with flexible and changeable goods size. The embodiment can customize the specification size of the pressure blanket 2, changes the common goods shelf into an intelligent goods shelf, avoids the waste of the conventional common goods shelf, and greatly saves the cost of additional purchasing goods shelf.

And S2, a pressure blanket setting step, namely setting a pressure blanket 2 on the upper surface of each tray 11, attaching the pressure blanket to the upper surface of each tray 11 and connecting the pressure blanket to the electronic equipment 20. The pressure blanket 2 is used for acquiring the weight of goods on the tray and the bottom surface contour thereof, and also is used for acquiring the contour of the tray area with the weight changed. The bottom surface profile of the article includes the shape and size of the bottom surface of the article. The pressure blanket 2 includes two blanket surfaces, and a plurality of pressure sensors 23 uniformly distributed between the two blanket surfaces and arranged in a matrix; wherein the distance between any two adjacent sensors is 2-10 mm, and the distance between two adjacent pressure sensors 23 is preferably 5mm. When the goods are placed on the pressure blanket 2 of one tray 11, the pressure sensor pressed by the goods generates an electric signal and sends the electric signal to the electronic device 20, and the electronic device 20 can calculate the contour of the area of the tray where the weight change occurs, that is, the contour of the bottom of the goods taken from the tray or placed on the tray, including the shape and size of the bottom of the goods, according to the electric signal transmitted by the pressure blanket 2. The electronic device 20 may also calculate the amount of change in weight of the tray, i.e., the weight of the item being removed from or placed on the tray.

S3, placing at least one article in at least one tray 11, preferably, each column is placed with the same article. The bottom of each goods is in direct contact with the pressure blanket on the upper surface of the tray, and the goods cannot be placed in a stacked mode, so that the bottom outline of each goods can be collected by the electronic equipment.

S4, a camera setting step, namely setting at least one camera 30 at the top of the unmanned convenience store and/or the front end of the goods shelf 10, connecting the at least one camera 30 to the electronic equipment 20, wherein the camera 30 is preferably a three-dimensional camera, and continuously acquiring multi-frame three-dimensional images. A part of cameras are arranged on the top of the convenience store, the field of view of the cameras covers the whole range of the bottom surface of the convenience store, and the electronic equipment judges the position of each user according to the real-time image of each user in the convenience store. When the weight of a certain tray changes, the electronic device judges whether the coordinate sets of the two hands of each user have an intersection with the coordinate set of the space above the tray, if so, the identity of the user stretching towards the tray when the weight of the tray changes is obtained, namely the identity of the user taking the goods from the tray or placing the goods to the tray is obtained. And a part of cameras are arranged at the front end of the goods shelf, and the vision field of the cameras covers the space in front of the goods shelf. When the weight of a certain tray changes, the electronic equipment intercepts an image of the space in front of the goods shelf in a time period before or after the time point, namely, a real-time image of the goods taken or put back by the user in a stretching mode, wherein the real-time image comprises all or part of pictures of the goods in the user, the pictures of the goods are input into a goods identification model by combining a computer vision technology and a machine learning technology, and the type of the taken or put back goods is further judged.

S5, an item database construction step, namely constructing an item database for storing a plurality of sets of item information of each item placed on the tray 11 of the goods shelf 10, wherein the item information comprises the type, the weight of each item, the column in which the item is placed and the bottom surface profile. As shown in fig. 7, the step S5 of constructing the item database specifically includes the steps of: s51, initializing an goods database, and inputting a goods shelf number, a tray number and a column number of each column on each tray; s52, inputting the goods information corresponding to each column to the goods database, wherein the goods information comprises the goods types corresponding to the column, the single-product weight and the price of the goods of the types; s53, at least one article is placed on a tray 11 of a goods shelf 10, and the bottom surface contour of each type of article is collected in real time and stored in the article database. Wherein the outline of the bottom surface of the goods comprises the shape of the bottom surface of the goods and the size of the bottom surface of the goods.

S6, a tray parameter acquisition step, namely acquiring weight change values of each tray 11 in real time; and when the weight change value is a non-zero value, recording tray parameters including the weight change value, a column with changed weight and a tray area contour with changed weight. Wherein the weight-changing tray area profile includes the shape and size of the weight-changing tray area.

When a user enters a store, the electronic equipment acquires the identity of each user entering the convenience store according to the two-dimensional code or the identity card of the user, and a shopping database is built for each user according to the identity information of the user. When the weight change value is a non-zero value, the electronic equipment acquires the hand position of each user in real time according to the camera at the height of the convenience store, and compares the hand position of the user with the position of the tray, so that a conclusion of which user stretches hands towards which tray can be obtained. When a shopping event occurs once, the weight of a certain column on the tray can change, and a user stretching to the column at the same time is a user taking or returning goods, so that the identity of the user taking or returning the goods can be judged, and the shopping database of the user can be synchronously updated.

S7, judging whether the weight change value X is smaller than 0 or not, if yes, judging that goods are taken away from a tray of a goods shelf, and executing the steps S8-S9; if not, it is determined that the article is placed on the tray of the shelf, and steps S10 to S14 are performed.

S8, a first goods judging step, namely judging that goods are taken from a certain tray of a goods shelf, judging the type, the quantity and the quantity of the taken goods according to the tray parameters and the goods information, and specifically comprising the following steps: comparing the number of the column with the number of the column of each article in the article database, obtaining the type of the article to be taken, and obtaining the weight of the single article of the article in the type from the article database; calculating the quotient of the absolute value of the weight change value and the weight of the single product of the type of the product in the product database, and carrying out rounding treatment on the quotient to obtain the number of the removed products.

S9, updating a shopping database of the user, storing the types and the quantity of the removed goods and the bottom surface profile of the removed goods to the shopping database of the user who removes the goods, recording the quantity change condition of each goods in the goods database, and updating the stock quantity of the goods on the goods shelves in the goods database.

S10, a second goods judging step, wherein goods are judged to be placed on a tray of the goods shelf, and the weight of the goods is the same as the weight change value.

S11, a weight comparison step, namely judging whether the weight change value, namely the weight of the goods, is the same as the weight of one goods or the total weight of a plurality of goods in a shopping database of a user placing the goods, if not, judging that the goods placed on a goods shelf are foreign matters; if so, M kinds of probability results of the kinds of the goods placed on the shelves are acquired, and the next step S12 is executed.

For example, if the item weighs 50 grams, there is only one 50 gram item in the user's shopping database, m=1. If the weight of the article is 200 grams, but there are two 200 grams of articles, three 100 grams of articles, one of the five articles in the shopping database of the user, the article placed on the shelf may be one 200 grams of article (two possible) or two 100 grams of article (three possible), and the total of five possible results, m=5. If the weight of the goods is 185 grams, the shopping database of the user has no 185 grams of goods, M=0, and the goods are judged to be foreign matters, such as a mobile phone of the user.

The foreign matter refers to other items not belonging to the convenience store that were not originally placed on the shelves of the convenience store. The electronic device may issue an alarm signal when the article type of an article is determined to be a foreign matter.

S12, a contour comparison step, namely judging whether the contour of a tray area with changed weight is consistent with the contour of at least one article in a shopping database of a user placing the article, if so, judging that the contour of the article placed on a goods shelf is identical with the contour of the article, and acquiring N kinds of possibility results of the type of the article placed on the goods shelf; if not, the article placed on the shelf is determined to be a foreign object by indicating that the weight of the article is the same but the contour is different from that of any article in the user shopping database.

For example, if the bottom profile of an item placed on a shelf is a square having a side length of 12 cm, only one item in the shopping database of the user has a square having a side length of 12 cm, and the weight of the item is the same as the weight of the item placed on the shelf, the item can be considered to be the item placed on the shelf. If the bottom profile of the goods placed on the shelf is a circle with a diameter of 10 cm, but the bottom profile of four goods in the shopping database of the user is a circle with a diameter of 10 cm, it can be considered that the goods may be one of the four goods, n=4. If the bottom contour of the goods placed on the goods shelf is a long bar with the length of 20 cm and the width of 3 cm, but the bottom contour of the goods in the shopping database of the user is not the same as the goods, n=0, the goods are judged to be foreign matters, and the goods can be carried by the user, such as an umbrella.

The sequence of steps S11 and S12 may be exchanged before and after, or performed synchronously, without mutual influence. In step S11 or S12, when m=0 or n=0, it may be determined that the placed article is a foreign matter.

And S13, judging the type of the goods placed back, wherein when M is more than 0 and N is more than 0, the type of the goods can be further determined by combining the possibility results obtained in the steps S11 and S12, and if the weight and the contour of the goods placed on the goods shelf are the same as those of one or more goods in the shopping database of the user, the goods placed on the goods shelf, namely the goods can be deduced. For example, it is known in step S11 that the weight of the article is 200 g, and in step S12 that the contour of the bottom surface of the article is a circle having a diameter of 10 cm; if only one of the five possible solutions of step S11 has a circular bottom contour with a diameter of 10 cm, or only one of the four possible solutions of step S12 has a weight of 200 g, the type and number of the returned articles can be determined.

In general, the total number of items purchased by a user in a convenience store is typically less than 10, and may occasionally reach 10 to 20. In the case of a relatively small sample size, the likelihood that the weight and profile of the goods placed on the shelf are the same as those of one or more goods in the shopping database of the user is substantially unique, and the type and number of the goods placed on the shelf can be determined by performing the above steps S11-13.

If the type and number of the goods placed on the shelf after the step S11-13 is executed are not unique, there are more than two possible results, and the difference between the possible results can be further identified by combining with other goods identification technologies, so as to accurately determine the type and number of the goods. Other item identification techniques include: the computer vision technology further judges the type of the goods by utilizing the real-time images of the goods; RFID identification technology, RFID tags are arranged on different goods, and the type of the goods is judged by using the RFID technology. S14, updating the shopping database of the user, and deleting the goods information corresponding to the goods from the shopping database of the user who returns the goods, so that the actual shopping situation of the user is consistent with the shopping database information. Updating an article database, and if the articles placed on the shelves are not foreign matters, storing the weight, the bottom surface profile, the types of the articles and the fields (namely the field numbers with the weight changed) of each article placed on the shelf as a group of article information into the article database.

Searching a column number corresponding to the goods from the goods database according to the types of the goods placed on the goods shelves, judging whether the column number corresponding to the types of the goods is consistent with the column number with changed weight, if so, judging that the goods are placed in the correct column, and only recording the quantity change condition of the goods database; if not, judging that the goods are placed in the wrong column, adding information of the misplaced goods in the goods information corresponding to the column in the goods database, and storing the goods types, the weights of single goods, the column and the bottom surface outline of each goods placed in the goods shelf as a group of goods information in the goods database. Optionally, the electronic device sends an alarm signal when an item is misplaced in the wrong field.

Steps S1 to S4 are all hardware setting steps, and steps S5 to S14 are all data processing steps executed by the electronic device 20.

As shown in fig. 8, the article identifying method of the present invention further includes a user identity acquiring step for acquiring the identity of the user who takes or places the article, including the following steps S15 to S17 performed by the electronic device 20. Steps S15 to S16 are performed before the tray parameter collection step S6, and step S17 is performed after the tray parameter collection step S6.

S15, user identity identification step, identify the identity information of each user. The user uses the mobile phone APP software to finish user registration, register identity information and a bank account for payment, and can generate a two-dimensional code by using the mobile phone when preparing to enter a store so as to indicate the identity. When a user enters a store, the electronic device acquires the two-dimensional code of the user according to the two-dimensional code reader at the gate of the store gate, so that the identity of each user entering the convenience store is acquired.

S16, generating a shopping database, namely generating the shopping database of each user according to the identity information of each user. The electronic device generates a shopping database for each user who is scanning into the store for recording shopping information of the user in the convenience store.

In the tray parameter collection step S6, if the weight change value of the tray is a non-zero number, the following step S17 is required to be executed after the step S6.

S17, a user identity judging step, namely judging the identity of the user taking or placing the goods, and updating a shopping database of the user. When the weight change value is smaller than 0, judging the identity of a user taking away the goods; and when the weight change value is larger than 0, judging the identity of a user placing the goods. When the weight change value is less than 0, in the following step S9, information of the user who removed the goods is recorded in the shopping database of the user. When the weight change value is greater than 0, invoking a shopping database of a user who places goods in the following steps S11 to S12; in the following step S14, information of the user placing the article is deleted from the shopping database of the user placing the article.

The invention installs the pressure blanket on the surface of the tray of the conventional goods shelf, and utilizes the pressure sensor matrix in the pressure blanket to collect the weight, position, bottom surface contour and other information of each goods placed on the pressure blanket in real time. When goods are taken from the goods shelf tray by a user, storing the weight and the bottom surface contour information of each taken goods into a shopping database of the user in real time; when a user places a certain article to a goods shelf tray, the shopping database of the user is called in time, the weight and the bottom surface outline of the article are compared with those of the shopping database of the user, and the weight and the bottom surface outline of the article and the shopping database are compared, so that whether the article is the article taken by the user before is judged, and the type and the number of the article are judged. In most cases, the electronic equipment can solve the problem of misplacement and misplacement by the method, and even if a user places the goods in the wrong column, the shopping record of the goods can be deleted from the shopping database of the user in time, so that the shopping database of the user is ensured to be consistent with the real shopping situation.

The invention has the beneficial effects that the weight and the bottom surface contour of each goods on the tray are acquired and recorded in a mode of installing the pressure blanket on the upper surface of the tray, and the tray area contour and the weight change value of the weight change on each tray are acquired in real time. When the goods are taken from a goods shelf tray or put back to the goods shelf tray, the type and the quantity of the goods can be timely and accurately judged, the shopping database of the user is updated in real time, the shopping record of the user in the shopping database is ensured to be consistent with the real shopping condition, the settlement accuracy of each user is ensured, the secondary sales of misplaced goods are not influenced, and the technical problem that the shopping database is wrong due to the fact that the user places the goods in an error column is effectively solved.

The goods shelf used in the invention has similar structure to the existing goods shelf, only a pressure blanket is needed to be installed on a common tray, the tray is not needed to be specially made into a key type structure with completely separated columns, and the cost of the goods shelf can be effectively reduced; the column width of each tray is adjustable, so that the tray can be used for placing goods with different sizes, is more in applicable scene, and can effectively reduce operation and management cost relative to the key type tray.

The foregoing is merely a preferred embodiment of the present invention to make a person skilled in the art more clearly understand how to practice the present invention, and these embodiments do not limit the scope of the present invention. Modifications and alterations will be apparent to those skilled in the art without departing from the principles of this invention, and it should be understood that this invention is not limited to the particular embodiments illustrated.

Claims

1. A method of identifying an article, comprising the steps of:

constructing an article database for storing article information of each article placed on a tray of a goods shelf, wherein the article information comprises the type, the weight of a single article, the column in which each article is positioned and the bottom surface contour;

collecting weight change values of each tray in real time; when the weight change value is a non-zero value, recording tray parameters including a weight change value, a column with a changed weight and a tray area contour with a changed weight; and

judging whether the weight change value is smaller than 0, if so, judging that the goods are taken from a tray of a goods shelf, judging the type and the number of the taken goods according to the tray parameters and the goods information, and storing the type, the number and the bottom surface outline of the taken goods into a shopping database of a user taking the goods; if not, judging that the goods are placed on the tray of the goods shelf, judging the types and the quantity of the goods placed on the goods shelf according to the tray parameters and the shopping database of the user placing the goods, and deleting the goods information corresponding to the goods from the shopping database of the user placing the goods;

a step of judging the type and number of goods placed on the shelf, comprising the steps of:

judging whether the weight of the goods is the same as the weight of at least one goods in a shopping database of a user placing the goods; if yes, obtaining M probability results of the types and the quantity of the goods placed on the goods shelf, namely weight comparison results;

judging whether the contour of the tray area with the weight changed is consistent with the contour of at least one article in a shopping database of a user placing the article; if yes, N possible results of the types and the quantity of the goods are obtained, namely a contour comparison result; and

and comparing the weight comparison result with the contour comparison result, and if one and only one of the weight comparison result and the contour comparison result are the same, obtaining the result of the type and the number of the goods placed on the goods shelf.

2. The article identification method according to claim 1, wherein,

judging the type and the number of the removed goods according to the tray parameters and the goods information, and specifically comprising the following steps:

comparing the number of the column with the number of the column in which each article in the article database is positioned, and obtaining the type of the article to be taken; and

calculating the quotient of the absolute value of the weight change value and the weight of the single product of the type of the product in the product database, and carrying out rounding treatment on the quotient to obtain the number of the removed products.

3. The article identification method as claimed in claim 1, further comprising the steps of:

identifying identity information of each user;

generating a shopping database of each user according to the identity information of each user; and

when the weight change value is smaller than 0, judging the identity of a user taking away the goods; and when the weight change value is larger than 0, judging the identity of a user placing the goods.

4. The article identification method according to claim 1, wherein,

the method for constructing the goods database specifically comprises the following steps:

initializing an goods database, and recording the column number of each tray into the goods database;

inputting the goods information corresponding to each column to the goods database, wherein the goods information comprises the types and the weights of the goods corresponding to the column; and

collecting the bottom surface profile of each article, and storing the bottom surface profile into the article database; when an article is placed on a tray, the contour of the area where the weight of the tray changes is the contour of the bottom surface of the article.

5. The article identification method as claimed in claim 1, further comprising the steps of:

deleting the item information of the taken item from the item database when the weight change value is less than 0;

when the weight change value is greater than 0, if the goods placed on the goods shelf are not foreign matters, the goods type, the single-product weight, the column and the bottom surface outline of each goods placed on the goods shelf are stored into the goods database as a group of goods information.

6. The article identification method as claimed in claim 1, further comprising the steps of:

searching a column corresponding to the goods from the goods database according to the type of the goods placed on the goods shelf;

judging whether the column corresponding to the goods of the category is consistent with the column with changed weight;

if yes, judging that the goods are placed in the correct columns;

if not, judging that the goods are placed in the wrong column, and optionally sending out an alarm signal.

7. The article identification method according to claim 1, wherein,

the bottom surface profile of the article includes the shape and size of the bottom surface of the article; and

the weight-changing tray area profile includes the shape and size of the weight-changing tray area.

8. The article identification method according to claim 1, wherein,

before the step of constructing the item database, the method further comprises the following steps of:

providing at least one shelf, each shelf comprising a plurality of trays, each tray being separated into at least one column;

a pressure blanket is arranged on the upper surface of each tray and is connected to an electronic device; and

placing at least one article in at least one tray, the bottom surface of each article contacting the top surface of the pressure blanket.

9. An electronic device, comprising:

a memory for storing executable program code; and

a processor connected to the memory for executing a computer program corresponding to the executable program code by reading the executable program code to perform the steps in the item identification method as claimed in any one of claims 1 to 6.

10. An item identification system, comprising:

the electronic device of claim 9;

at least one shelf, each shelf comprising a plurality of trays, each tray being divided into a plurality of columns for placing goods; and

at least one pressure blanket attached to the upper surface of each tray and connected to an electronic device; the pressure blanket is used for acquiring weight change values of the tray, columns with weight change and outlines of the tray area with weight change in real time.

11. The item identification system of claim 10 wherein,

the pressure blanket includes a plurality of pressure sensors evenly distributed.

12. The item identification system of claim 10, further comprising:

and the isolating piece is detachably arranged above each tray and is used for separating the tray into a plurality of columns.