CN109979130A - A kind of commodity automatic identification and clearing sales counter, method and system - Google Patents
A kind of commodity automatic identification and clearing sales counter, method and system Download PDFInfo
- Publication number
- CN109979130A CN109979130A CN201910247358.7A CN201910247358A CN109979130A CN 109979130 A CN109979130 A CN 109979130A CN 201910247358 A CN201910247358 A CN 201910247358A CN 109979130 A CN109979130 A CN 109979130A
- Authority
- CN
- China
- Prior art keywords
- commodity
- obstacle sensor
- sales counter
- video image
- sensor state
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07F—COIN-FREED OR LIKE APPARATUS
- G07F9/00—Details other than those peculiar to special kinds or types of apparatus
- G07F9/002—Vending machines being part of a centrally controlled network of vending machines
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07G—REGISTERING THE RECEIPT OF CASH, VALUABLES, OR TOKENS
- G07G1/00—Cash registers
- G07G1/12—Cash registers electronically operated
- G07G1/14—Systems including one or more distant stations co-operating with a central processing unit
- G07G1/145—PLU-management
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
- H04N7/181—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast for receiving images from a plurality of remote sources
Abstract
The invention discloses a kind of commodity automatic identifications and clearing sales counter, method and system, are related to automatic identification field.Video image, detection obstacle sensor state and the changed timestamp of the detection obstacle sensor state when this method is purchased by obtaining commodity；According to the timestamp of detection obstacle sensor state and detection obstacle sensor state change, the video image is segmented；The type of merchandise and commodity amount in each section of video image are identified using commodity identification model, each segment mark of the video image is shown as commodity taking-up section or commodity put back to section, and obtains the quantity purchase of all kinds of commodity；According to the unit price and quantity purchase of all kinds of commodity, goods amount is calculated.Technical solution of the present invention can reduce the usage amount of camera, improve sales counter space utilization rate, and support commodity are stacked to put back to random layer, reduce consuming cost, and increase the flexibility of merchandising.
Description
Technical field
The present invention relates to automatic identification field, in particular to a kind of commodity automatic identification and clearing sales counter, method and
System.
Background technique
With the arriving in business transformation epoch, the retail trade in transition forward position obtains the extensive pass of Liao Ge large enterprises
Note.The old management mode of traditional retail mode is hard to carry in the increasingly developed current era of computer vision technique, and new zero
The epoch of selling come into being.
Epoch product of the intelligent sales counter as new retail trade, commodity identification similarly live through several research cycles:
Original adoption RFID Radio Frequency Identification Technology, the program need to stick corresponding label for each commodity, not only expend label cost,
It also needs to consume a large amount of costs of labor, the unrecognized problem of commodity caused by falling off there is also RFID label tag.Known based on image
Other technology multiple fields successful application, it is also natural that certain Research foundation is provided for retail trade.Before research
Phase, before being opened by still image identification technology to cabinet door and cabinet door close after commodity amount in captured two pictures into
Row compares, and obtains the taking-up classification and corresponding number of commodity.This method recognition effect is preferable, but needs to install for each layer of shelf
One camera not only increases camera cost；In order to guarantee shooting angle, every layer of counter will reserve certain space,
Sales counter space utilization rate is wasted significantly；And static identification technology cannot support commodity to stack, and commodity is not supported to change layer and put yet
It returns.
Summary of the invention
In order to overcome technical problem as described above, the present invention completes dynamic using a kind of commodity dynamic recognition technique and takes object
Commodity identification in the process, takes out commodity in conjunction with detection obstacle sensor and judges with the state of putting back to, be finally completed certainly
Dynamic clearing.The invention can reduce the usage amount of camera, improve sales counter space utilization rate.
Specific technical solution of the present invention is as follows:
In a first aspect, the present invention proposes a kind of commodity automatic identifying method, comprising:
Obtain video image, detection obstacle sensor state and the detection barrier sensing when commodity are purchased
When device state is changed
Between stab；
According to the timestamp of detection obstacle sensor state and detection obstacle sensor state change, to the view
Frequency image is segmented；
The type of merchandise and commodity amount in each section of video image are identified using commodity identification model, by the video
Each segment mark of image is shown as commodity taking-up section or commodity put back to section；
Section is taken out according to commodity and commodity put back to the type of merchandise of section and commodity amount counts the commodity respectively and increases
With the quantity of reduction, the quantity purchase of all kinds of commodity is obtained.
Second aspect, a kind of commodity automatic settlement method, comprising:
S1 obtains video image, detection obstacle sensor state and the detection barrier when commodity are purchased
The changed timestamp of sensor states；
S2, according to the timestamp of detection obstacle sensor state and detection obstacle sensor state change, to institute
Video image is stated to be segmented；
S3 identifies the type of merchandise and commodity amount in each section of video image using commodity identification model, will be described
Each segment mark of video image is shown as commodity taking-up section or commodity put back to section；
S4 takes out section according to commodity and commodity put back to the type of merchandise of section and commodity amount counts the commodity respectively and increases
Reduced quantity is summed it up, the quantity purchase of all kinds of commodity is obtained；
S5 calculates goods amount according to the unit price and quantity purchase of all kinds of commodity.
Further, before the S3, comprising:
S31 obtains the image of all kinds of commodity as sample data；
The sample data is divided into training data and test data according to preset ratio by S32；
S33 establishes commodity identification model by the training data based on neural network structure；
S34 tests the commodity identification model by the test data, and according to test result to the quotient
Product identification model is adjusted.
Further, further includes:
S6 deducts the quotient from account bound in the user id automatically according to the corresponding user id of video image
The product amount of money, and generate corresponding order information.
The third aspect, the present invention proposes a kind of commodity Automatic-settlement remote server, including processor and memory, described
Memory is stored with an at least Duan Chengxu, and described program is executed by the processor to realize following method:
Receive user's request；
Send door opening command；
Receive video image, detection obstacle sensor state and the detection barrier sensing when commodity are purchased
The changed timestamp of device state；
Receive door signal；
Commodity are identified and settled accounts using the commodity automatic settlement method as described in first aspect is any.
Fourth aspect, the present invention propose a kind of commodity Automatic-settlement sales counter, and detection barrier is equipped with around the sales counter
Hinder object sensor and camera, the sales counter is used for equipped with local server, the local server:
Receive door opening command；
Cabinet door of selling goods is opened in control；
Control the camera starting video capture；
The detection changed timestamp of obstacle sensor state described in detection obstacle sensor state is read in real time；
Upload video image captured by the camera, the detection obstacle sensor state and detection barrier
Hinder the changed timestamp of object sensor states；
When cabinet door of selling goods is closed, door signal is sent, and triggers the camera and stops video capture.
Further, the detection obstacle sensor is grating sensor, and quantity is 2, is respectively arranged in sales counter
Two sides up and down or the left and right sides.
Further, the camera is 2, is respectively arranged in the top of sales counter, and be located remotely from each other.
5th aspect, the present invention propose a kind of commodity Automated Clearing House system, comprising:
Sales counter and remote server are equipped with detection obstacle sensor and camera around the sales counter, described
Sales counter is attached between the local server and remote server by communication network equipped with local server, institute
Stating communication network includes cable network or wireless network,
The local server, is used for:
Receive door opening command transmitted by the remote server；
Cabinet door of selling goods is opened in control；
Control the camera starting video capture；
When reading detection obstacle sensor state and the changed detection obstacle sensor state in real time
Between stab；
Video image, the detection obstacle that the camera is shot when commodity are purchased are uploaded to the remote server
Object sensor states and the changed timestamp of the detection obstacle sensor state；
When cabinet door of selling goods is closed, Xiang Suoshu remote server sends door signal, and triggers the camera and stop view
Frequency is shot；
The remote server, is used for:
Receive user's request；
Door opening command is sent to the local server；
Receive video image, detection barrier that the local server uploads camera shooting when commodity are purchased
Hinder object sensor states and the changed timestamp of the detection obstacle sensor state；
Receive door signal transmitted by the local service；
Commodity are identified and settled accounts using the commodity automatic settlement method as described in first aspect is any.
Further, the detection obstacle sensor is grating sensor, and quantity is 2, is respectively arranged in sales counter
Two sides up and down or the left and right sides.
Further, the camera is 2, is respectively arranged in the top of sales counter, and be located remotely from each other.
Technical solution provided by the invention has the benefit that
Video image, detection obstacle sensor state and the detection when present invention is purchased by obtaining commodity
The changed timestamp of obstacle sensor state；According to detection obstacle sensor state and detection obstacle sensor
The timestamp of state change is segmented the video image；It is identified in each section of video image using commodity identification model
The type of merchandise and commodity amount, each segment mark of the video image be shown as commodity take out section or commodity putting back to section；According to
Commodity take out section and commodity put back to the type of merchandise of section and commodity amount counts the quantity that the commodity increase and decrease respectively,
Obtain the quantity purchase of all kinds of commodity；According to the unit price and quantity purchase of all kinds of commodity, goods amount is calculated.With the prior art
It compares, technical solution of the present invention has the advantages that
(1) usage amount of camera can be reduced；
(2) sales counter space utilization rate is improved, and supports commodity are stacked to put back to random layer, is reduced to a certain extent
Consuming cost, and increase the flexibility of merchandising.
Detailed description of the invention
In order to illustrate the technical solution of the embodiments of the present invention more clearly, required use in being described below to embodiment
Attached drawing be briefly described, it should be apparent that, drawings in the following description are only some embodiments of the invention, for this
For the those of ordinary skill of field, without creative efforts, it can also be obtained according to these attached drawings others
Attached drawing.
Fig. 1 show a kind of commodity automatic identification proposed by the present invention and settlement system schematic diagram；
Fig. 2 show a kind of commodity automatic identification of the present invention and settlement method schematic diagram；
Fig. 3 show a kind of building schematic diagram of commodity identification model of the present invention；
Fig. 4 is another commodity automatic identification of the present invention and settlement method schematic diagram；
Fig. 5 shows a kind of commodity automatic identification involved in the embodiment of the present invention and the structure of clearing remote server is shown
It is intended to.
Specific embodiment
To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with attached drawing to embodiment party of the present invention
Case is described in further detail.
The present invention discloses a kind of commodity automatic identifying method, comprising:
Obtain video image, detection obstacle sensor state and the detection barrier sensing when commodity are purchased
When device state is changed
Between stab；
According to the timestamp of detection obstacle sensor state and detection obstacle sensor state change, to the view
Frequency image is segmented；
The type of merchandise and commodity amount in each section of video image are identified using commodity identification model, by the video
Each segment mark of image is shown as commodity taking-up section or commodity put back to section；
Section is taken out according to commodity and commodity put back to the type of merchandise of section and commodity amount counts the commodity respectively and increases
With the quantity of reduction, the quantity purchase of all kinds of commodity is obtained.
This commodity automatic identifying method can be applied to different application scenarios, for example apply this approach to commodity
In settlement system, in actual operation, the detection obstacle sensor optionally can be radar, it is preferred that be light
Gate sensor.It is excellent with grating sensor when being illustrated to technical solution of the present invention in the description of following description
Embodiment is selected to be described, but technical solution of the present invention is not limited thereto.
It is as shown in Figure 1 a kind of commodity automatic identification proposed by the present invention and settlement system schematic diagram, comprising:
Sales counter 100 and remote server 110 are equipped with grating sensor 101 and camera around the sales counter
102, the sales counter 100 passes through between the local server 103 and remote server 110 equipped with local server 103
Communication network is attached, and the communication network includes cable network or wireless network.
The Goodsshelves of sales counter 100 in this system can stack, and can save material-putting space.Sales counter 100 can be
Refrigerator-freezer.In possible realization, grating sensor 101 is mounted in pairs, and local server 103 receives between two sensor regions
Interior detects the state change situation for whether having barrier, for the subsequent taking-up for judging commodity or puts back to state, it is preferred that
A pair of of grating sensor is respectively arranged in the two sides up and down or the left and right sides of sales counter, is to be able to recognize each layer of shelf
The process for taking object and putting back to.
In possible practical operation, camera 102 is installed on the top of sales counter, it is preferred that 2 cameras, it is optional
, it is high definition wide-angle camera, is respectively arranged in the top of sales counter, and be located remotely from each other, to maximum magnitude takes and sell
The video image that commodity are removed or put back in counter.
Sales counter 100 also has the display screen that can be used for showing information, when cabinet door is in close state, the display screen institute
The information of display includes the two dimensional code acquired in the remote server 110, and user can carry out it by Third-party payment software
Scanning optionally can be common at present wechat or Alipay software.It should be noted that in possible practical operation,
User carries out cabinet door unlatching using the dynamic two-dimension code generated in display screen above the payment softwares such as Alipay/wechat scanning refrigerator-freezer
When, if the user used for the first time, then exempt from close payment function firstly the need of opening small amount, barcode scanning carries out cabinet door unlatching again.
When user wants the commodity in the purchase sales counter 100, pass through Third-party payment software scans sales counter 100
On two dimensional code, remote server 110 receive user request, user request information here may include that the user is registered
Third-party payment account information.
In a kind of possible realization, remote server 110 judges the user for new user according to user request information
When, new mark can be generated for the user automatically, be user id optionally uniquely to distinguish over other users, when judging the use
When family is not new user, the information that can uniquely indicate the user is read, is user id optionally.
After remote server 110 receives user's request, door opening command can be sent to the local server 103, and
After local server 103 receives door opening command transmitted by the remote server 110, cabinet door of selling goods is opened in control, simultaneously
The camera starting video capture is controlled, and uploads camera when commodity are purchased to the remote server 110 and claps
The video image taken the photograph, in a kind of possible realization, by the way of real-time uploaded videos image, in alternatively possible realization
In, by the way of timing uploaded videos image.
After the cabinet door of sales counter 100 is opened, local server 103 reads grating sensor state, and recording light in real time
The timestamp of gate sensor state change.Preferably, grating sensor here can mounted in pairs in the two sides up and down of sales counter
Or the left and right sides, it changes so that user when taking out or putting back to commodity, can trigger grating sensor state.
Remote server 110 receives the view that the local server 103 uploads camera shooting when commodity are purchased
Frequency image, the grating sensor state and the changed timestamp of grating sensor state.
After user chooses the commodity of all desired purchases from sales counter 100, cabinet door can be actively closed, sales counter is worked as
When door is closed, local server 103 can send door signal to the remote server, and trigger the camera and stop video
Shooting.
After remote server 110 receives door signal transmitted by the local service 103, it will start according to from originally
The commodity that ground server 103 the is uploaded camera 102 is shot when purchased video image, the grating sensor 101
The timestamp of 101 state change of state and grating sensor is identified and is settled accounts commodity.It is illustrated in figure 2 the present invention one
Kind commodity automatic identification and settlement method schematic diagram, show the specific implementation step of this method, comprising:
In step 201, video image, grating sensor state and the grating sensing when commodity are purchased are obtained
The changed timestamp of device state；
Video image when being purchased according to commodity can identify the type of merchandise and commodity number by image recognition technology
Amount；According to grating sensor state and the changed timestamp of grating sensor state, can obtain in sales counter
Commodity period for being removed or putting back to.It should be noted that grating sensor here is at least a pair, sales counter cabinet door
It is opened, user stretches out one's hand by during thing, and grating sensor state can change.Specifically, the light being oppositely arranged
If being obstructed between gate sensor without barrier, then communicating between grating sensor, state is communicating state；If
When having barrier between the two, the state of grating sensor is obstructed state at this time, such as is sold goods here when user reaches into
When taking out in cabinet or putting back to commodity, the state of grating sensor can be made to become obstructed state from communicating state.Therefore, in this step
In rapid, obtain video image, grating sensor state and the grating sensor state when commodity are purchased and change
Timestamp as in subsequent step carry out commodity identification and clearing data source.
In step 202, according to the timestamp of grating sensor state and grating sensor state change, to the view
Frequency image is segmented；
According to the timestamp of grating sensor state and grating sensor state change, in conjunction with step 201, i.e. grating is passed
Sensor is become the timestamp of obstructed state from communicating state, is denoted as T1；And become the timestamp of communicating state from obstructed state,
It is denoted as T2.So, the T1 moment be may be to reach up at the time of reach into sales counter, it is also possible to the last time is taken out
Commodity put back to sales counter；The T2 moment is that hand is taken out the moment from sales counter, may be that object is taken to go out sales counter, it is also possible to by quotient
At the time of product are put back to hand removal sales counter after sales counter.And this period of T1-T2, grating sensor are constantly in obstructed shape
State, i.e., select the period of commodity in sales counter, the data of this time can not consider.Before directly considering each T1 moment
Some time corresponding video image, put back to section as doubtful commodity；And some time after each T2 moment is corresponding
Video image, as doubtful commodity take out section.It should be noted that above-mentioned doubtful commodity put back to section and doubtful commodity take out section
Need to carry out subsequent further image recognition also to complete accurate commodity identification.
In step 203, the type of merchandise and commodity number in each section of video image are identified using commodity identification model
Each segment mark of the video image is shown as commodity taking-up section or commodity puts back to section by amount；
It has completed to be segmented the video image by above-mentioned steps 202, every section of all corresponding goods are removed or put
The process of returning.This step will further indicate each section of the video image, be denoted as commodity taking-up section or commodity are put
Return section.Video image has recorded the dynamic image that user changes over time when buying commodity, so it is easy to understand that only relies on view
Frequency image is removed or puts back to judge commodity, is very difficult, and only identifies that article corresponding to image is
It is relatively simple.In a kind of possible realization, for taking out initial image frame and end picture frame, knot in each section of video image
Step 202 is closed, in a kind of possible practical operation, corresponding initial image frame here is grating sensor by communicating state
Image under becoming captured by some time before the timestamp of obstructed state, corresponding end picture frame is grating sensor
Captured lower image after becoming the timestamp of communicating state from obstructed state, by comparing initial frame and end frame image
If having no commodity in initial frame image, and there are commodity in the frame image of end in possible realization in commodity situation of change, that
, corresponding video image section, signable is that commodity take out section；If there is commodity in initial frame image, and nothing in the frame image of end
Commodity, then, corresponding video image section is signable to put back to section for commodity.
It should be noted that initial image frame and end picture frame among the above does not need centainly to correspond to correct time
Stamp can select multiframe to carry out comprehensive descision, in the allowed band of certain time difference more accurately to identify the commodity of image
Type and commodity amount.
Commodity identification model in this step just can export the corresponding type of merchandise and commodity number by input picture frame
Amount.
In step 204, section is taken out according to commodity and commodity put back to the type of merchandise of section and commodity amount counts respectively
The quantity that the commodity increase and decrease obtains the quantity purchase of all kinds of commodity；
In this step, the commodity amount being removed and the commodity amount being put back into are counted respectively, for calculating commodity
For quantity purchase, the former is increased quantity, and the latter is reduced quantity.
In step 205, according to the unit price and quantity purchase of all kinds of commodity, goods amount is calculated.
The type of merchandise recognized by step 201 to step 204 inquires its corresponding cargo price, and multiplied by phase
Quantity purchase is answered, then different types of commodity price is summed, just calculates the goods amount that user needs to pay.
In a kind of possible practical operation, it is illustrated in figure 3 a kind of building schematic diagram of commodity identification model of the present invention,
Show a kind of building implementation process of commodity identification model described in step 203 in embodiment corresponding to Fig. 2, comprising:
In step 301, the image of all kinds of commodity is obtained as sample data；
In possible realization, the high-definition image of sales counter institute vending articles is shot indoors, in the insufficient feelings of data volume
Under condition, data enhancing can also be completed by modes such as overturning, rotation and random croppings.In alternatively possible realization, also
The picture frame about commodity can be obtained by simulating true shopping video.
In step 302, the sample data is divided into training data and test data according to preset ratio；
General training data volume can be greater than amount of test data.
In step 303, commodity identification model is established by the training data based on neural network structure；
In possible practical operation, in the yolov3 network models under original darknet frame, pass through step
301 and step 302 sales counter commodity training data obtained, carry out the training of commodity identification model.
In step 304, the commodity identification model is tested by the test data, and according to test result
The commodity identification model is adjusted.
In possible realization, by step 303, after the training of commodity identification model is completed, using test data to mould
Type is tested, test data here, optionally, be can be true shopping picture frame, is also possible to captured original image
The process image of picture or original image after treatment.According to the convergence state of test result analysis "current" model, if model is complete
The problem of the problem of not restraining, detection is needed to be training data or network structure.It is not received if model caused by wrong data
It holds back, needs to acquire correct data again and be trained；The problem of if network structure, then needs adjustment network depth, non-linear journey
Degree etc..If partial convergence occurs in model, analysis test effect is model over-fitting or poor fitting.It, can if model over-fitting
By cleaning data again, increasing amount of training data, network complexity, addition Dropout are reduced, learning rate or reduction are reduced
The number of epoch is finely tuned；If model poor fitting, then need by increasing network complexity, increasing learning rate, reduction regularization
Parameter is increased network nonlinearity and is finely adjusted using modes such as batch normalization.If model is received completely
It holds back, then can keep every time other ginsengs in the parameters such as learning rate, batchsize, epoch, filter scales, filter quantity
Number is constant, and the mode for only adjusting one of parameter carries out small parameter perturbations.
If Fig. 4 is another commodity automatic identification of the present invention and settlement method schematic diagram, comprising:
In step 401, video image, grating sensor state and the grating sensing when commodity are purchased are obtained
The changed timestamp of device state；
In step 402, according to the timestamp of grating sensor state and grating sensor state change, to the view
Frequency image is segmented；
In step 403, the type of merchandise and commodity number in each section of video image are identified using commodity identification model
Each segment mark of the video image is shown as commodity taking-up section or commodity puts back to section by amount；
In step 404, section is taken out according to commodity and commodity put back to the type of merchandise of section and commodity amount counts respectively
The quantity that the commodity increase and decrease obtains the quantity purchase of all kinds of commodity；
In step 405, according to the unit price and quantity purchase of all kinds of commodity, goods amount is calculated.
Above-mentioned steps 401 are similar with the corresponding steps in embodiment corresponding to Fig. 2 to step 405, and details are not described herein again.
In a step 406, it according to the corresponding user id of video image, is detained from account bound in the user id automatically
Except the goods amount, and generate corresponding order information.
In possible realization, by calling Third-party payment interface, institute in account bound in user id is deducted in request
Goods amount is stated, realizes auto deduction.
Video image, grating sensor state and the grating sensing when the present embodiment is purchased by obtaining commodity
The changed timestamp of device state；According to the timestamp of grating sensor state and grating sensor state change, to institute
Video image is stated to be segmented；The type of merchandise and commodity number in each section of video image are identified using commodity identification model
Each segment mark of the video image is shown as commodity taking-up section or commodity puts back to section by amount；Section is taken out according to commodity and commodity are put back to
The type of merchandise and commodity amount of section count the quantity that the commodity increase and decrease respectively, obtain the purchase number of all kinds of commodity
Amount；According to the unit price and quantity purchase of all kinds of commodity, goods amount is calculated.Compared with prior art, technical solution of the present invention
It has the advantages that
(1) usage amount of camera can be reduced；
(2) sales counter space utilization rate is improved, and supports commodity are stacked to put back to random layer, is reduced to a certain extent
Consuming cost, and increase the flexibility of merchandising.
Fig. 5 shows a kind of commodity automatic identification involved in the embodiment of the present invention and the structure of clearing remote server is shown
It is intended to, the device mainly includes processor 501, memory 502 and bus 503, the memory is stored with an at least Duan Chengxu,
Described program is executed by the processor to realize a kind of following commodity automatic identifications and settlement method applied to sales counter:
Receive user's request；
Send door opening command；
Video image, grating sensor state and the grating sensor state when commodity are purchased is received to become
The timestamp of change；
Receive door signal；
Using as described in above method embodiment commodity automatic identification and settlement method identified and settled accounts commodity.
Processor 501 includes one or more processing cores, and processor 501 passes through bus 503 and 502 phase of memory
Even, memory 502 realizes above-mentioned application when executing the program instruction in memory 502 for storing program instruction, processor 501
In the commodity automatic identification and settlement method of sales counter.
Optionally, memory 502 can be by any kind of volatibility or non-volatile memory device or their group
It closes and realizes, such as static to access memory (SRAM) at any time, electrically erasable programmable read-only memory (EEPROM) is erasable to compile
Journey read-only memory (EPROM), programmable read only memory (PROM), read-only memory (ROM), magnetic memory, flash
Device, disk or CD.
The present invention also provides a kind of computer readable storage medium, be stored in the storage medium at least one instruction,
At least a Duan Chengxu, code set or instruction set, at least one instruction, an at least Duan Chengxu, code set or instruction set are by institute
Processor is stated to load and execute to realize above-mentioned commodity automatic identification and settlement method applied to sales counter.
Optionally, the present invention also provides a kind of computer program products comprising instruction, when it runs on computers
When, so that computer executes above-mentioned commodity automatic identification and settlement method applied to sales counter.
Those of ordinary skill in the art will appreciate that realizing that all or part of the steps of above-described embodiment can pass through hardware
It completes, relevant hardware can also be instructed to complete by program, the program can store computer-readable with one kind
In storage medium, storage medium mentioned above can be read-only memory, disk or CD etc..
The present invention discloses a kind of commodity automatic identification and clearing sales counter, is equipped with grating sensor around the sales counter
And camera, the sales counter are used for equipped with local server, the local server:
Receive door opening command；
Cabinet door of selling goods is opened in control；
Control the camera starting video capture；
Grating sensor state is read in real time, and records the timestamp of grating sensor state change；
Upload video image captured by the camera, the grating sensor state and the grating sensor shape
The changed timestamp of state；
When cabinet door of selling goods is closed, door signal is sent, and triggers the camera and stops video capture.
Preferably, the grating sensor is 2, is respectively arranged in the two sides up and down or the left and right sides of sales counter.
Preferably, the camera is 2, is respectively arranged in the top of sales counter, and be located remotely from each other.
Commodity automatic identification and clearing sales counter in the embodiment of the present invention, on the basis of realizing automatic vending commodity,
Allow shelf to stack, saves space, meanwhile, the camera quantity to identify commodity is less, economizes on resources.
The foregoing is merely presently preferred embodiments of the present invention, is not used to limit invention, it is all in spirit of the invention and
Within principle, any modification, equivalent replacement, improvement and so on be should all be included in the protection scope of the present invention.
Claims (11)
1. a kind of commodity automatic identifying method characterized by comprising
Obtain video image, detection obstacle sensor state and the detection obstacle sensor shape when commodity are purchased
The changed timestamp of state；
According to the timestamp of detection obstacle sensor state and detection obstacle sensor state change, to the video figure
As being segmented；
The type of merchandise and commodity amount in each section of video image are identified using commodity identification model, by the video image
Each segment mark be shown as commodity and take out section or commodity putting back to section；
Section is taken out according to commodity and commodity put back to the type of merchandise of section and commodity amount counts the commodity respectively and increases and subtract
Few quantity obtains the quantity purchase of all kinds of commodity.
2. a kind of commodity automatic settlement method characterized by comprising
S1 obtains video image, detection obstacle sensor state and the detection barrier sensing when commodity are purchased
The changed timestamp of device state；
S2, according to the timestamp of detection obstacle sensor state and detection obstacle sensor state change, to the view
Frequency image is segmented；
S3 identifies the type of merchandise and commodity amount in each section of video image using commodity identification model, by the video
Each segment mark of image is shown as commodity taking-up section or commodity put back to section；
S4, according to commodity take out section and commodity put back to the type of merchandise of section and commodity amount count respectively the commodity increase and
The quantity of reduction obtains the quantity purchase of all kinds of commodity；
S5 calculates goods amount according to the unit price and quantity purchase of all kinds of commodity.
3. commodity automatic settlement method according to claim 2, which is characterized in that before the S3, comprising:
S31 obtains the image of all kinds of commodity as sample data；
The sample data is divided into training data and test data according to preset ratio by S32；
S33 establishes commodity identification model by the training data based on neural network structure；
S34 tests the commodity identification model by the test data, and is known according to test result to the commodity
Other model is adjusted.
4. commodity automatic settlement method according to claim 2, which is characterized in that further include:
S6 deducts the commodity gold according to the corresponding user id of video image from account bound in the user id automatically
Volume, and generate corresponding order information.
5. a kind of commodity Automatic-settlement remote server, which is characterized in that including processor and memory, the memory storage
There is an at least Duan Chengxu, described program is executed by the processor to realize following method:
Receive user's request；
Send door opening command；
Receive video image, detection obstacle sensor state and the detection obstacle sensor shape when commodity are purchased
The changed timestamp of state；
Receive door signal；
Commodity are identified and settled accounts using the commodity automatic settlement method as described in claim 2 to 4 is any.
6. a kind of commodity Automatic-settlement sales counter, which is characterized in that be equipped with detection obstacle sensor around the sales counter
And camera, the sales counter are used for equipped with local server, the local server:
Receive door opening command；
Cabinet door of selling goods is opened in control；
Control the camera starting video capture；
The detection changed timestamp of obstacle sensor state described in detection obstacle sensor state is read in real time；
Upload video image captured by the camera, the detection obstacle sensor state and the detection barrier
The changed timestamp of sensor states；
When cabinet door of selling goods is closed, door signal is sent, and triggers the camera and stops video capture.
7. commodity Automatic-settlement sales counter according to claim 6, which is characterized in that the detection obstacle sensor is
Grating sensor, quantity are 2, are respectively arranged in the two sides up and down or the left and right sides of sales counter.
8. commodity Automatic-settlement sales counter according to claim 6 or 7, which is characterized in that the camera is 2, point
It is not installed on the top of sales counter, and is located remotely from each other.
9. a kind of commodity Automated Clearing House system characterized by comprising
Sales counter and remote server are equipped with detection obstacle sensor and camera around the sales counter, described to sell goods
Cabinet is attached between the local server and remote server by communication network equipped with local server, described logical
Communication network includes cable network or wireless network,
The local server, is used for:
Receive door opening command transmitted by the remote server；
Cabinet door of selling goods is opened in control；
Control the camera starting video capture；
Detection obstacle sensor state and the changed timestamp of the detection obstacle sensor state are read in real time；
The camera is shot when commodity are purchased video image, the detection barrier is uploaded to the remote server to pass
Sensor state and the changed timestamp of the detection obstacle sensor state；
When cabinet door of selling goods is closed, Xiang Suoshu remote server sends door signal, and triggers the camera and stop video bat
It takes the photograph；
The remote server, is used for:
Receive user's request；
Door opening command is sent to the local server；
Receive video image, the detection barrier that the local server uploads camera shooting when commodity are purchased
Sensor states and the changed timestamp of the detection obstacle sensor state；
Receive door signal transmitted by the local service；
Commodity are identified and settled accounts using commodity automatic settlement method as claimed in claim 2.
10. commodity Automated Clearing House system according to claim 9, which is characterized in that the detection obstacle sensor is
Grating sensor, quantity are 2, are respectively arranged in the two sides up and down or the left and right sides of sales counter.
11. commodity Automated Clearing House system according to claim 9 or 10, which is characterized in that the camera is 2, point
It is not installed on the top of sales counter, and is located remotely from each other.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910247358.7A CN109979130A (en) | 2019-03-29 | 2019-03-29 | A kind of commodity automatic identification and clearing sales counter, method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910247358.7A CN109979130A (en) | 2019-03-29 | 2019-03-29 | A kind of commodity automatic identification and clearing sales counter, method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109979130A true CN109979130A (en) | 2019-07-05 |
Family
ID=67081519
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910247358.7A Pending CN109979130A (en) | 2019-03-29 | 2019-03-29 | A kind of commodity automatic identification and clearing sales counter, method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109979130A (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110458110A (en) * | 2019-08-14 | 2019-11-15 | 青岛英泰软件技术有限公司 | It is capable of the retail method of automatic identification fruit variety |
| CN110570283A (en) * | 2019-09-11 | 2019-12-13 | 炫盛(上海)科技有限公司 | shopping method and system based on block chain |
| CN110782200A (en) * | 2019-09-10 | 2020-02-11 | 成都亿盟恒信科技有限公司 | Intelligent management system and method for logistics vehicles |
| CN110991273A (en) * | 2019-11-18 | 2020-04-10 | 支付宝(杭州)信息技术有限公司 | Laminate regulating and controlling equipment, method and device |
| CN111507315A (en) * | 2020-06-15 | 2020-08-07 | 杭州海康威视数字技术股份有限公司 | Article picking and placing event detection method, device and equipment |
| CN111626201A (en) * | 2020-05-26 | 2020-09-04 | 创新奇智(西安)科技有限公司 | Commodity detection method and device and readable storage medium |
| CN111815852A (en) * | 2020-07-07 | 2020-10-23 | 武汉马克到家科技有限公司 | Image and gravity dual-mode automatic commodity identification system for open-door self-taking type sales counter |
| JP2021093649A (en) * | 2019-12-11 | 2021-06-17 | キヤノンマーケティングジャパン株式会社 | Information processing apparatus, method for controlling information processing apparatus, and program |
| CN113178032A (en) * | 2021-03-03 | 2021-07-27 | 北京迈格威科技有限公司 | Video processing method, system and storage medium |
| CN113191176A (en) * | 2020-06-08 | 2021-07-30 | 阿里巴巴集团控股有限公司 | Data processing method and device |
| CN113194252A (en) * | 2021-04-28 | 2021-07-30 | 深圳市乐科智控科技有限公司 | Data flow consumption optimization method and system for dynamically identifying intelligent cabinet |
| CN113469644A (en) * | 2021-06-18 | 2021-10-01 | 深圳市点购电子商务控股股份有限公司 | Shelf life reminding method, automatic vending device, computer equipment and storage medium |
| CN113506400A (en) * | 2021-07-05 | 2021-10-15 | 深圳市点购电子商务控股股份有限公司 | Automatic vending method, automatic vending device, computer equipment and storage medium |
| CN113727029A (en) * | 2021-11-03 | 2021-11-30 | 武汉星巡智能科技有限公司 | Intelligent order generation method for combining collected images at multiple visual angles and intelligent vending machine |
| CN114333146A (en) * | 2021-12-31 | 2022-04-12 | 元气森林(北京)食品科技集团有限公司 | Display cabinet, display cabinet information acquisition method and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108921645A (en) * | 2018-06-07 | 2018-11-30 | 深圳码隆科技有限公司 | A kind of commodity purchasing determination method, device and user terminal |
| CN108921098A (en) * | 2018-07-03 | 2018-11-30 | 百度在线网络技术（北京）有限公司 | Human motion analysis method, apparatus, equipment and storage medium |
| CN108985359A (en) * | 2018-06-29 | 2018-12-11 | 深圳和而泰数据资源与云技术有限公司 | A kind of commodity recognition method, self-service machine and computer readable storage medium |
| CN109147170A (en) * | 2018-07-26 | 2019-01-04 | 上海凯景信息技术有限公司 | A kind of self-service cabinet based on image recognition |
| US20190034901A1 (en) * | 2017-07-28 | 2019-01-31 | Convipay Inc. | Self service terminal enabling auto-pressing of keys and method thereof |
| CN109447619A (en) * | 2018-09-20 | 2019-03-08 | 华侨大学 | Unmanned settlement method, device, equipment and system based on open environment |
-
2019
- 2019-03-29 CN CN201910247358.7A patent/CN109979130A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190034901A1 (en) * | 2017-07-28 | 2019-01-31 | Convipay Inc. | Self service terminal enabling auto-pressing of keys and method thereof |
| CN108921645A (en) * | 2018-06-07 | 2018-11-30 | 深圳码隆科技有限公司 | A kind of commodity purchasing determination method, device and user terminal |
| CN108985359A (en) * | 2018-06-29 | 2018-12-11 | 深圳和而泰数据资源与云技术有限公司 | A kind of commodity recognition method, self-service machine and computer readable storage medium |
| CN108921098A (en) * | 2018-07-03 | 2018-11-30 | 百度在线网络技术（北京）有限公司 | Human motion analysis method, apparatus, equipment and storage medium |
| CN109147170A (en) * | 2018-07-26 | 2019-01-04 | 上海凯景信息技术有限公司 | A kind of self-service cabinet based on image recognition |
| CN109447619A (en) * | 2018-09-20 | 2019-03-08 | 华侨大学 | Unmanned settlement method, device, equipment and system based on open environment |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110458110A (en) * | 2019-08-14 | 2019-11-15 | 青岛英泰软件技术有限公司 | It is capable of the retail method of automatic identification fruit variety |
| CN110782200A (en) * | 2019-09-10 | 2020-02-11 | 成都亿盟恒信科技有限公司 | Intelligent management system and method for logistics vehicles |
| CN110570283A (en) * | 2019-09-11 | 2019-12-13 | 炫盛(上海)科技有限公司 | shopping method and system based on block chain |
| CN110991273A (en) * | 2019-11-18 | 2020-04-10 | 支付宝(杭州)信息技术有限公司 | Laminate regulating and controlling equipment, method and device |
| JP2021093649A (en) * | 2019-12-11 | 2021-06-17 | キヤノンマーケティングジャパン株式会社 | Information processing apparatus, method for controlling information processing apparatus, and program |
| CN111626201B (en) * | 2020-05-26 | 2023-04-28 | 创新奇智(西安)科技有限公司 | Commodity detection method, commodity detection device and readable storage medium |
| CN111626201A (en) * | 2020-05-26 | 2020-09-04 | 创新奇智(西安)科技有限公司 | Commodity detection method and device and readable storage medium |
| CN113191176A (en) * | 2020-06-08 | 2021-07-30 | 阿里巴巴集团控股有限公司 | Data processing method and device |
| CN111507315A (en) * | 2020-06-15 | 2020-08-07 | 杭州海康威视数字技术股份有限公司 | Article picking and placing event detection method, device and equipment |
| CN111815852A (en) * | 2020-07-07 | 2020-10-23 | 武汉马克到家科技有限公司 | Image and gravity dual-mode automatic commodity identification system for open-door self-taking type sales counter |
| CN113178032A (en) * | 2021-03-03 | 2021-07-27 | 北京迈格威科技有限公司 | Video processing method, system and storage medium |
| CN113194252A (en) * | 2021-04-28 | 2021-07-30 | 深圳市乐科智控科技有限公司 | Data flow consumption optimization method and system for dynamically identifying intelligent cabinet |
| CN113469644A (en) * | 2021-06-18 | 2021-10-01 | 深圳市点购电子商务控股股份有限公司 | Shelf life reminding method, automatic vending device, computer equipment and storage medium |
| CN113506400A (en) * | 2021-07-05 | 2021-10-15 | 深圳市点购电子商务控股股份有限公司 | Automatic vending method, automatic vending device, computer equipment and storage medium |
| CN113727029A (en) * | 2021-11-03 | 2021-11-30 | 武汉星巡智能科技有限公司 | Intelligent order generation method for combining collected images at multiple visual angles and intelligent vending machine |
| CN114640797A (en) * | 2021-11-03 | 2022-06-17 | 深圳友朋智能商业科技有限公司 | Order generation method and device for synchronously optimizing commodity track and intelligent vending machine |
| CN114333146A (en) * | 2021-12-31 | 2022-04-12 | 元气森林(北京)食品科技集团有限公司 | Display cabinet, display cabinet information acquisition method and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109979130A (en) | A kind of commodity automatic identification and clearing sales counter, method and system | |
| US10192408B2 (en) | Registry verification for a mechanized store using radio frequency tags | |
| CN108922026B (en) | Replenishment management method and device for vending machine and user terminal | |
| WO2020047919A1 (en) | Self-service vending method, apparatus and system, and server and computer-readable storage medium | |
| WO2020010827A1 (en) | Open type vending machine-based shopping settlement method and device | |
| CN111263224B (en) | Video processing method and device and electronic equipment | |
| CN108198315A (en) | A kind of auth method and authentication means | |
| CN109711827A (en) | A kind of new retail method of payment and payment system of near-eye display device | |
| CN108876504A (en) | A kind of unmanned selling system and its control method | |
| CN111222870B (en) | Settlement method, device and system | |
| US11023728B1 (en) | Machine learning algorithm trained to identify algorithmically populated shopping carts as candidates for verification | |
| US20210398097A1 (en) | Method, a device and a system for checkout | |
| CN110490697A (en) | Unmanned convenience store's settlement method, device, computer and storage medium | |
| Sarwar et al. | Smart shopping carts based on mobile computing and deep learning cloud services | |
| CN110717769A (en) | Goods returning method, device, equipment and storage medium for intelligent cabinet articles | |
| CN109523694A (en) | A kind of retail trade system and method based on commodity detection | |
| CN111178860A (en) | Settlement method, device, equipment and storage medium for unmanned convenience store | |
| CN113468914B (en) | Method, device and equipment for determining purity of commodity | |
| US20220309784A1 (en) | System and method for populating a virtual shopping cart based on a verification of algorithmic determinations of items selected during a shopping session in a physical store | |
| CN109190706A (en) | Self-service method, apparatus and system | |
| CN110689389A (en) | Computer vision-based shopping list automatic maintenance method and device, storage medium and terminal | |
| CN111260685B (en) | Video processing method and device and electronic equipment | |
| CN108764410B (en) | Unmanned store goods returning method and device and electronic equipment | |
| CN110490698A (en) | A kind of unmanned misplaced detection method and device of store shelf article | |
| US20230005348A1 (en) | Fraud detection system and method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190705 |