CN105303204A - Quick article identification - Google Patents

Abstract

A method and a device are provided for quick article identification. A plurality of sensors for unknown articles passing through the sensors capture data. The data are processed to generate a plurality of physical attributes related with the unknown articles. The physical attributes are used for searching for physical attributes of a large amount of known articles in a database, wherein the unknown articles are one of the known articles. A small group of the known articles are selected, and the physical attributes of the known articles are consistent with the physical attributes of the unknown articles; and by further processing the selected group of the known articles, the unknown articles are identified as one group of the known articles.

Description

Rapid item identification

Technical field

The present invention relates to improving equipment and method of article identification, particularly relate to the equipment for article shown in point of sales terminal identification and method.

Background technology

It is existing for many years that retail shop uses the point of sales terminal being equipped with optical code scanner to process purchase-transaction.During this period, utilize optical code scanner to carry out swindling the method bought to emerge in an endless stream.Swindle method comprises with identifying that the optical code of low value article replaces the optical code identifying high-value items.Swindle method also comprises accommodation trade or article is delivered to bagging station and does not scan.The bagging station that the trial that minimizing is swindled is included in point of sales terminal sets up safe scale, guarantees that the article packed conform to the purchase of identification.But these trials do not eliminate swindle.For above-mentioned situation, if being equal in weight of low value article and high-value items, so safe scale is swindled with regard to None-identified, because can weigh up predetermined weight at bagging station.

In addition, some problems irrelevant with swindle are also had.Such as, the optical code of identify objects because damaging, being hidden by other labels or the marking, or may lack and cannot read.In these cases, people must be had to identify article and information is inputted point of sales terminal.Usually, this can increase the time needed for purchase-transaction.

For addressing these problems, now write out without optical code, only just can the software of accurate identify objects with outward appearance.This software carrys out identify objects by analyzing the images of items obtained.But, analyze the time that the image obtained needs relatively to grow, and analyze the data record very huge (sometimes close to 100,000 byte) produced.Analyze and also need the huge database including all known contents, for accurately identifying unknown article.In some embodiments, database purchase exceedes 100, the data of 000 kind of unknown article, and the data-base recording of often kind of known contents has 10,000 to 100,000 byte.Final database needs the data space of 1 to 10G.If be placed in by database on central server computer, so to the article that each part will identify, all need the data transmitting 10,000 to 100,000 byte between point of sales terminal and central server computer.The data of article are sent to central server computer, and then carrying out processing can the several seconds consuming time with the process of identify objects.Only have the database of 80 known contents data to test with one, the average recognition time drawn is 2 to 3 seconds.For having 100, the database of 000 article, estimated average recognition time is 3,651 seconds.

Summary of the invention

In its various aspects, the present invention is intended at least one the shortcoming overcoming or improve prior art, or provides a kind of useful replacement scheme.

According to a first aspect of the invention, the invention provides a kind of computer manner of execution, utilize the physical attribute identification of article for the article of item scanner identification, the method comprises: use one or more sensors of item scanner, to calculate in the multiple relational physical properties of article the numerical value of each; From multiple physical attribute, identify the physical attribute having predefine classification, wherein all predefine classifications are all based on the preset range of physical property values, and identified for all the predefine classification that physical attribute determines to comprise calculated value; Generate the synthesized attribute data record comprising each physical attribute of multiple physical attribute; Multiple synthesized attribute data records of the known contents of the synthesized attribute data record generated and storage are compared; And choose one group of known contents data record stored, in this record, the synthesized attribute data record that stores of known contents data record selected by each conforms to the synthesized attribute data record generated.

In addition, method in claim 1, wherein eachly store the part that synthesized attribute data record is known contents data record, and stored the physical attribute that synthesized attribute data record also represents known contents, and each data record that stored also comprises information for identifying known contents.

In addition, multiple physical attribute comprises metal properties, containing the instruction that there is metal in article.

In addition, multiple physical attribute comprises temperature property, the instruction containing article temperature.

In addition, temperature instruction comprises the pre-set categories such as freezing, refrigeration, room temperature and high temperature.

In addition, multiple physical attribute comprises hardness property, the instruction containing article rigidity.

In addition, hardness instruction comprises the pre-set categories such as soft, half soft and hard.

In addition, multiple physical attribute comprises color configuration attribute, the instruction containing 8 kinds of Color Channels.

In addition, multiple sensor comprises the detectable sensor that whether there is metal.

In addition, comprise can every the sensor of empty detecting article temperature for multiple sensor.

In addition, multiple sensor comprises multiple video camera, is carried out by the image of picked-up processing to calculate shape attribute, size attribute and color configuration attribute.

In addition, the synthesized attribute data record size generated is equal to or less than 8 data bytes.

According to a second aspect of the invention, the invention provides an item scanner, it comprises: shell; The item transport of article can be transmitted via shell; And being positioned at the non-contact temperature sensor of shell, this sensor can measure article by temperature during shell.

This item scanner also can comprise metal detection sensor for whether there is metal in detecting article.

Above-mentioned shell also at least can comprise a tower, and each tower comprises multiple video camera, these cameras capture to image can calculate shape attribute, size attribute and color configuration attribute through process.

This item scanner also can comprise the platform scale in a shell.

This item scanner also can comprise a ultrasonic transmitter-receiver, for the hardness property of detecting article.

This item scanner also can comprise a programmed processor, and the information that this processor uses one or more sensor to catch calculates each property value of the multiple physical attributes relevant to article.

Shell defines one in addition and receives the input door of article and withdraw from the output door of article, and all the sensors is all positioned at shell.

Above-mentioned programmed processor measures shape attribute, size attribute and color configuration attribute separately by multiple video camera.

According to a third aspect of the invention we, the invention provides an item scanner and identify article for its scanning, this item scanner comprises: one comprises the entrance receiving article and the shell of outlet exiting article; Item transport article being reached outlet from the entrance of shell; Be positioned at multiple sensors of shell, when article are by obtaining its information during shell; A storer comprising computer instruction; One with the processor of memory communication, when processor performs computer instruction, processor can be caused to be implemented as follows function: calculate multiple physical attributes value separately relevant to article by the information that the one or more sensors in multiple sensor obtain, wherein a part for multiple physical attribute comprises predefined type, and the value of these physical attributes also comprises a class predefined type; And choose data to identify each cover known contents, wherein relevant to known contents multiple physical attributes value separately all conforms to multiple physical attributes value separately of being correlated with these article.

In addition, above-mentioned multiple physical attribute comprises weight attribute, and wherein multiple sensor comprises a platform scale, for determining the weight of the article of treated device process, to obtain weight attribute.

In addition, above-mentioned multiple physical attribute also comprises shape attribute, size attribute and color configuration attribute, wherein multiple sensor comprises multiple video camera, for catching the image (the treated device process of image) of article, to obtain shape, size and color configuration property value.

In addition, above-mentioned color configuration attribute comprises data 8 kinds of Color Channels.

In addition, above-mentioned multiple physical attribute comprises metal properties, and multiple sensor comprises metal detector, for determining whether there is metal in article, and generates the instruction after treated device process, to obtain the value of metal properties.

In addition, the plurality of physical attribute comprises temperature property, wherein multiple sensor comprises non-contact temperature sensor, for determine apart from the position of article certain distances article temperature (the treated device process of temperature of these article, can from one group comprise likely temperature range predetermined temperature classification choose a class).

In addition, the plurality of physical attribute also comprises hardness property, and wherein multiple sensor comprises ultrasonic transmitter-receiver, for determining the hardness number (choosing a class from one group of predetermined hardness classification) of the article of treated device process.

According to a fourth aspect of the present invention, the invention provides the storage instruction of one or more permanent digital storage media, this instruction, when one or more computer equipment performs, can cause method listed in first aspect present invention to be implemented.

According to a fifth aspect of the present invention, the invention provides as identify objects and create the method for reduced data collection, the method comprises: the multiple physical attributes (comprising size, weight and tenor) measuring article; Generate multiple through measuring physical attribute synthesized attribute data record separately; The synthesized attribute data record generated and known contents multiple are stored compared with synthesized attribute data record; And choose one group of known contents data record stored, wherein each selected known contents data record the synthesized attribute of storage data record with generate synthesized attribute data record and conform to.

Therefore, with regard to its each side, the present invention shows to carry out identify objects by article outward appearance, and only relies on minute quantity data to identify that each article are desirable.Only carry out identify objects with low volume data and decrease the data volume needing to store and search for, also shorten the time needed for identify objects.Meanwhile, because optical code is not used in identify objects, therefore rely on outward appearance to carry out identify objects and decrease swindle.

In its each side, the present invention shows by article identification process being divided into two independences and different functions can reduce time needed for identify objects.Section 1 function eliminates a large amount of possibility identification items of article, may mate and reduce to less identification collection.Such as, this vast number suitably may comprise 100, more than 000 record, and each record is loaded with the information of identifiable design Individual Items.Section 1 function adopts the attribute of unknown article to be reduced to by this vast number suitably may comprise the less collection recorded less than 10.Littlely any one is concentrated to record the unknown article of all identifiable design.This function needs to carry out a large amount of process.Section 2 function uses the method being different from Section 1 function to identify the unknown article remained in a small amount of article record.

In its each side, the present invention shows further, and the minority physical attribute relevant with article can be used for will to be reduced to possibility identification item on a small quantity by identification item in a large number.With regard to each article to be identified, these attributes can measured and process within the relatively short time (usual less than 1 second).Such as, use this method, the quantity of possible identification item can be down to less than 10 from reaching 100,000 rapidly, only need use a database needing to store less than 10 megabytes of data.

The article of an item scanner identification can reach 100 altogether, 000 more than one piece.In its each side, the present invention uses the synthesized attribute that just unknown article and known contents generate that article total amount is reduced to several the article matched with unknown goods attribute.Once article total amount reduces to a small group candidate item, this group candidate item can be used to perform extra process, thus determine the characteristic of unknown article.In some cases, this group candidate item only has one, therefore just can identify unknown article without the need to extra process.

According to an embodiment of the invention, the invention provides one and implemented by item scanner, to identify the method for article.The method comprises: use the sensor of one or more item scanner to calculate the value of each physical attribute of the multiple physical attributes relevant to article; From multiple physical attribute, identify the physical attribute having predefine classification, wherein all predefine classifications are all based on the preset range of physical property values, and identified for all the predefine classification that physical attribute determines to comprise calculated value; Generate the synthesized attribute data record comprising each physical attribute of multiple physical attribute; Multiple synthesized attribute data records that stored of the synthesized attribute data record generated and known contents are compared; And choose one group of known contents data record stored, in this record, the synthesized attribute data record that stores of known contents data record selected by each conforms to the synthesized attribute data record generated.

Accompanying drawing explanation

Below to describe in detail and drawing of enclosing will contribute to comprehend the present invention and other features and advantages of the present invention.

The present invention may present with various assembly and arrangements of components and various method.Drawing is only demonstrated as the explanation of a few class embodiment and alternative method, must not explain limitation of the present invention.Drawing is not necessarily drawn in proportion.In each drawing, similar elements numbering is for describing the same parts in each drawing, picture and chart.

Fig. 1 is high level block diagram, represents the point of sales system set up according to the present invention.

Fig. 2 A is high level block diagram, represents the selected assembly of scanner/scale.

Fig. 2 B is high level block diagram, represents the selected assembly of an article identification shell of scanner/scale.

Fig. 3 is a form, lists the physical attribute of the article measured by scanner/scale.

Fig. 4 is high level flow chart, represents the method for operating with scanner/scale identification article.

Embodiment

All kinds of details have been listed, to help to understand the invention of advocating in below describing.But these those skilled in the art it should be understood that some aspect of institute's claimed invention may be implemented when not applying all following details, and possibility and expectation can make all kinds of change or amendment to described embodiment.

See Fig. 1, this is a secondary high level block diagram, represents point of sales system 100.The point of sales system 100 of present embodiment comprises the point of sales terminal 110 adopting network 155 to communicate with a storage server computer 160.One or more article that the plan that point of sales terminal 110 is shown by identification client is bought carry out purchase-transaction.In purchase transaction process, point of sales terminal 110 communicates with storage server computer 160, to send and to receive the data relevant with this purchase-transaction.

Although only have a point of sales terminal 110 to be shown in figure, it should be noted that this point of sales system 100 supports multiple point of sales terminals 110 that use network 155 communicates with storage server computer 160.In addition, it should be noted that point of sales terminal 110 suitably can be presented as the point of sales terminal of an auxiliary point of sales terminal or client's operation.In some embodiments, point of sales system 100 comprises the point of sales terminal of auxiliary point of sales terminal and client's operation.

Network 155 suitably can comprise a network, and this Web vector graphic is based on the communication protocol of transmission control protocol/Internet protocol (TCP/IP).Network 155 suitably can comprise the combination of LAN (Local Area Network) and wide area network.Network 155 suitably can comprise the combination in any of wireless network and wired network further.Wireless network comprises computer wireless LAN (Local Area Network) and the data network based on honeycomb.

Storage server computer 160 comprises a processor and comprises the storer of application software.The processor of storage server computer 160 performs this application software, and this application software makes this processor implement to support characteristic and the function of the operation of this storer.The characteristic that this application software provides and function may suitably comprise: to the support of point of sale operation, articles seeking database, sale and stock control, personal management and client help service.

Point of sales terminal 110 suitably comprises a computer 115, scanner/scale device 120, magnetic stripe reader/PIN (Personal Identification Number) input equipment 125, key board unit 130, cashing machine/spot assets module and tells paper money device 135, printer apparatus 140, operator's display 145 and a computer network 150.Computer network 150 can comprise a more than class network, and each network is for communicating with distinct device.Computer 115 is by computer network 150 and devices communicating, and computer network 150 suitably can comprise the enforcement of the USB (universal serial bus) (USB) of industry standard.Computer network 150 additionally can comprise second network for communicating with display device (as operator's display 145).

Computer 115 suitably can comprise the personal computer for using in another equipment (as point of sales terminal 110).In some embodiments, computer 115 is single-borad computers.Computer 115 comprises a processor, storer, network controller (for controlling external network 155) and a computer network controller (for controlling computer network 150).The storer of computer 115 comprises the computer instruction performed by the processor of computer 115, and these instructions can allow processor control assembly and the equipment of point of sales terminal 110, and provide the Premium Features of point of sales terminal 110.In some embodiments, storer comprises the instruction of a database and this database of certain operations.

Magnetic stripe reader/PIN (Personal Identification Number) input equipment 125 is a magnetic stripe reader and PIN (Personal Identification Number) (PIN) equipment.It reads the magnetic strip information through the card of this equipment.Such as, device 125 reads credit card, debit card and the magnetic stripe at the purchase card back side and the magnetic stripe at some driving license back sides.PIN (Personal Identification Number) input equipment allows client or operator to input may PIN (Personal Identification Number) associated with this card.This information subsequently can by computer network 150 being transferred on computer by safety.

Cashing machine/spot assets module and tell paper money device 135 and suitably can comprise a cashing machine equipment or a spot assets module and tell paper money device equipment, or comprise both simultaneously.Operator's supplementary embodiments of point of sales terminal 110 suitably can comprise this cashing machine equipment, this is because have office worker to operate this currency.The embodiment of client's operation of point of sales terminal 110 suitably can comprise spot assets module and tell paper money device equipment, and it can protect currency, but allows client give and receive currency.But in other embodiments, cashing machine equipment, spot assets module and tell paper money device equipment and all exist.This cashing machine/spot assets module and tell paper money device equipment 135 and use computer network 150 communicate with computer 115 and control by it.

Operator's display 145 comprises an electronic equipment, and this electronic equipment shows to client or operator the information received from computer 115 by computer network 150.Operator's display 145 also comprises the touch-screen input device of detection time of being touched of touch screen and position, and sends this information to computer 115 by computer network 150.Some embodiments have a more than operator's display 145, and in these embodiments, an operator's display 145 is used by the office worker of operation point of sales terminal 110, and second operator's display 145 is used by the client done shopping.

Refer to now Fig. 2 A and 2B.Fig. 2 A is high level block diagram, represents the selected assembly of scanner/scale (scanner) 120.Fig. 2 B is high level block diagram, represents the selected assembly of an article identification shell (shell) 280 of scanner 120.Scanner 120 identifies the article that the plan identified for it is bought, and this process is a part for the purchase-transaction that point of sales terminal 110 operates.Scanner 120 uses multiple sensor 270, and these sensors may be positioned at shell 280, to measure or to determine the multiple physical attributes relevant with the article 290 through shell 280.

Scanner 120 suitably can comprise a processor 205, interface circuit 210, storer 215, user interface 265 and multiple sensor 270.Multiple sensor 270 suitably can comprise a platform scale 235, multiple cameras 240, ultrasonic transmitter-receiver 245, item transport 250, metal detector 255 and a temperature sensor 260.Interface circuit 210 provides electronic equipment, and processor 205 needs by these electronic equipments, the miscellaneous equipment of usage data network 230 and storer 215, computer network 150 and scanner 120 and component communication.Interface circuit 210 produces data network 230, and this data network can suitably comprise electronic equipment and software, to produce the USB (universal serial bus) (USB) of industry standard.

Storer 215 comprises computer instruction 220, and these instructions are performed by processor 205, makes processor 205 perform characteristic and the function of scanner 120.When processor 205 performs computer instruction 220, cause equipment and the assembly of processor 205 scanner 120 further.In some embodiments, storer 215 also comprises a database 225, and computer instruction 220 comprises some instructions like this: when these instructions are performed by processor 205, will cause characteristic and the function of processor 205 usage data storehouse 225 performing database.

In the present embodiment, scanner 120 comprises article identification shell 280, and this shell comprises an entrance 281 (receiving article to be used for identifying) and outlet 282 (article exit shell 280 at this).In other embodiments, shell 280 has other physical form, such as carousel.In addition, the position of distinct device in shell 280 should not be construed as fixing.Although the position drawn can use, only for explanation.Item transport 250 receives article at the entrance 281 of shell 280, and article are transmitted in the inside 284 via shell 280, are then sent via outlet 282 by article.The operation of item transport 250 is controlled by processor 205.In some embodiments, item transport 250 is made up of multiple independently item transport, these independently item transport to work in coordination mobile article.

Scanner 120 uses multiple sensor 270 to measure or determines the specific physical attribute of each article by shell 280.The output of each sensor comprises the particular data of the type sensor.Such as, platform scale 235 exports its gravimetric value born at special time, and video camera 240 produces image, and metal detector 255 produces at short notice from the data that the delivery outlet of the detecting coil of metal detector 255 is caught.The data that sensor produces are further processed, to obtain available attribute data.Such as, the data received from metal detector 255 represent a reflected signal and must process, to determine whether to detect metal.

Metal detector 255 is positioned at shell 280, on its detecting article forwarder 250 through the article of metal detector 255 be metal or non-metal article.An article (such as tank vegetables) can be identified as metal, because it is packaged in metal can.But, will non-metal article be identified as with the refrigerated vegetables of nonmetallic plastic bag packaging.For detecting the existence of metal, metal detector 255 suitably can use pulse induction technology or ultra-low frequency (inductance balance) technology or other suitable technology.

Platform scale 235 is positioned at shell 280, and measures the weight through the article of shell 280.In some embodiments, item transport 250 is installed in the LOAD CELLS of platform scale 235.LOAD CELLS measures the weight of item transport 250 and upper article thereof subsequently.Then, platform scale 235 deducts tare weight from measured weight, thus determines the weight of article.In other embodiments, item transport 250 is divided into three independently Individual Items forwarders.First item transport brings article into shell 280 by entrance 281, and by these article stored in second item transport.Second item transport is positioned at shell 280 and is supported by the LOAD CELLS of platform scale 235.This LOAD CELLS measures the weight of these article and second item transport.Platform scale 235 deducts tare weight from measured weight, thus determines the weight of the article on second item transport.Once determine the weight of article, article move to by second item transport and stored in the 3rd item transport, article are shifted out shell 280 by outlet 282 by the 3rd item transport.

Temperature sensor 260 is for measuring the temperature of article when article pass shell 280.In some embodiments, temperature sensor 260 comprises the infrared non-contact temperature sensor that is measured article temperature from afar.In other embodiments, temperature sensor 260 is far infrared sensor arraies, and it can determine more large regions but not the temperature of single-point.Temperature sensor 260 can the temperature of the suitably article of measuring tempeature scope between 0 °F to 200 °F.The temperature recorded, for determining whether these article are frozen as frozen ice-cream, is refrigerated as milk, as a box cereal, be in room temperature, or is hot as the roast chicken bought from delicatessen.Such as, one records the article that temperature is equal to or less than 30 °F and is confirmed as freezing.One records temperature higher than 30 °F but the article being equal to or less than about 55 °F are confirmed as refrigeration.One records temperature higher than 55 °F but under the article being equal to or less than 90 °F are confirmed as being in room temperature.One records temperature and is confirmed as being hot higher than the article of 90 °F.It should be noted that the temperature range for determining article temperature state can regulate as required.Such as, in the open-air atmosphere that the market of farm produce is such, in 24 hours, the temperature of surrounding environment may change between 40 °F to 100 °F, therefore needs ceaselessly to regulate temperature range.Technology will be measured and correspondingly adjusting ambient temperature.

Ultrasonic transmitter-receiver 245 is for determining the hardness of article when article pass shell 280.When article are near ultrasonic transmitter-receiver 245, ultrasonic transmitter-receiver 245 is launched ultrasonic signal and is also measured the signal reflected from article to these article.This measures the hardness for determining these article.The hardness recorded is divided into three predetermined classifications: 1) hard; 2) half is soft; And 3) soft.Each classification can cause article to reflect the ultrasonic energy of different amount, and the ultrasonic energy of different amount produces different measurement reactions.The article of hard can reflect maximum energy, and soft article then reflect minimum energy.The energy of half soft article reflection is between therebetween.In other embodiments, air feeder is used to replace ultrasonic transmitter-receiver 245.Air feeder penetrates one of short duration weak gas flow to article, then measures the signal of reflection air pressure.Different hardness produces different reflection air pressure signals.In other embodiments, when article are moved to one known surface (as metal surface), a voice signal can be captured.Voice signal is used to the hardness determining article.One tank cotton candy and one bag of cotton candy drop to known surperficial time will have different voice signals, and this voice signal determines their hardness.

Video camera 240 comprises multiple cameras or image-capturing apparatus.Shell 280 comprises two towers 285 being positioned at item transport 250 offside.Each tower can suitably comprise multiple cameras 240.Video camera 240 in each tower 285 aims at the center of item transport 250.In some embodiments, one or two tower 285 extends to the top of item transport 250, and video camera 240 is just arranged on here, so as to overlook and catch through the image of article.When article between two towers 285 through out-of-date, video camera 240 can catch the view data of these article at least both sides.In addition, tower 285 is provided with light fixture to illuminate the article through tower 285.This light fixture suitably can comprise LED, and the light quantity produced by this equipment is controlled by processor 205.Tower 285 can suitably comprise towards the reference marker of item transport 250 side.Visible in the reference marker of each tower video camera 240 over there on tower, and in the image that caught of being captured.Reference marker can suitably for determining the size and dimension of article.

In some embodiments, the position of sensor 270 and article all different through the order of sensor 270.In other embodiments, do not use shell 280, sensor 270 is in open state, but item transport 250 still promotes object through sensor 270.

User interface 265 is used to communicate with the personnel operating scanner 120.User interface 265 can suitably comprise one or more lamp and/or loudspeaker.Processor 205 will pass on some information by operation light, and pass on some information that can be heard by the user by speakers generate sound.These sound may suitably comprise the human language copied by processor 205.

Fig. 3 provides a table 300, and this table lists the physical attribute of the article measured by scanner 120 or calculated.Table 300 lists seven physical attributes that attribute column 305 is specified.Scanner 120 uses the one or more sensors 270 shown in Fig. 2 B to measure, calculate or determine seven physical attributes 305.Each sensor 270 produces raw data usually, and this raw data needs realize the measurement to one of them attribute in seven physical attributes by processor 205 extra process.Measure the probable value that row 310 list each physical attribute, or to the description of measuring.Figure place row 315 define to be needed to store the computer data position representing each physical attribute metrical information.

First physical attribute is " metallicity ".Metal properties is used to indicate is existed metal by the article 290 that scan.Scanner 120 uses metal detector 255 sensor to carry out scan articles 290.When article 290 are through metal detector 255, processor 205 receives and processes the raw data from metal detector 255, to determine whether there is metal.The measurement 310 of metal properties has two predetermined classifications: metal or nonmetal.Owing to only having two kinds of possible classifications, a data bit is only needed to store this metal properties.In the example described in Fig. 2 B, are tank bean or pea by the article 290 scanned.This jar is made of metal, and therefore when article 290 are through metal detector 255, metal detector 255 output display exists the data of metal.Processor 205 receives and exports data, and after these data of process, determine that article 290 are containing metallics.Then the value of metal properties is set as " metal " by processor 205.If example article 290 is polybag bean or pea, metal detector 255 can not detect metal, so the value of metal properties is set as " nonmetal " by processor 205.

Second physical attribute is " shape ".Shape attribute defines the general physical form of article 290.Shape attribute has four predetermined classifications: square, cylindrical, oval and other shapes.Shape attribute is that the one or more seizure images by processing article 290 are determined.Wherein a video camera 240 catches image, and the view data of processor 205 to one or more image processes, to determine that it meets the general geometric configuration of one of predetermine class.If this shape is not mated with square, cylindrical or ellipse, then select " other shapes " classification.Shape attribute needs two data bit to represent four kinds of possible predetermine class.

3rd physical attribute is " hardness ".Hardness property identification is used for the material of connection with wrapping of piece 290 or forms the hardness of article 290 assembly.Hardness property has three predetermine class: soft, half soft and hard.By using ultrasonic transmitter-receiver 245 sensor to send ultrasound wave to article 290, and measure the sound wave of reflection from article 290, thus measure its hardness property.Processor 205 processes the data representing and record sound wave, to determine the classification representing article 290 hardness in three predetermine class.Because hardness property only has three predetermine class, therefore only need the classification that two data bit express possibility.

In other embodiments, air feeder is used to the hardness determining article 290.Air feeder sends one Tiny pore assembled facing to article 290, and the barometric wave then by measuring reflection determines the hardness of article 290.Article through measuring are harder, and the pressure wave amplitude of reflection is larger.

4th physical attribute is " temperature ".Temperature property measures the Current Temperatures of article 290, and is attributed to one of four predetermine class.This classification is divided into freezing, refrigeration, room temperature and high temperature.Scanner 120 serviceability temperature sensor 260 reads the temperature of article 290 when temperature sensor 260.Temperature sensor 260 is a non-contact infrared temperature sensor preferably, the infrared energy that sensing article 290 are launched, to determine the surface temperature of article 290.Temperature sensor 260 exports data, represents the surface temperature that article 290 record.Processor 205 receives temperature data and processes these data further, to determine that this temperature belongs to any of four kinds.Such as, freezing classification is by any temperature be defined as in advance lower than 30 °F.Therefore, if the temperature recorded is 25 °F, processor 205 will select freezing classification as its temperature property.Because temperature property only has four classifications, therefore only need the classification that two data bit express possibility.

5th physical attribute is " weight ".Weight attribute refers to the weight measuring article 290.Scanner 120 uses platform scale 235 to determine the actual weight of article 290.Processor 205 receives the actual weight being accurate to several decimals from platform scale 235.Processor 205 converts actual weight to ounce, and ounce is multiplied by two, is then rounded up to immediate integer.Such as, the weight of 1 pound 1/2 ounce (about 467.8 grams) is multiplied by 2 after being converted to 16.5 ounces and obtains 33.The 33 increment numbers representing 1/2 ounce (about 14.2 grams) in 1 pound 1/2 ounce.10 data bit are used to store weight attribute.This means that the highest weight that can be stored is 32 pounds (about 14.5 kilograms).This attribute does not have the classification defined in advance, but actual weight is converted into self-defined unit, and resolution is lower than actual weight, and maximal value is 32.Extra figure place can be added process larger maximum weight value through confirmation, or incremental weight value is increased to 1 ounce (about 28.4 grams) or more, to process larger maximum weight value from 1/2 ounce (about 14.2 grams).Other unit (as SI unit) can be used for alternative English unit.

6th physical attribute is " size ".Size attribute refers to that the three-dimensional (length) of the physical size of article 290 is measured.Scanner 120 catches by process article 290 the size that image determines article 290.This image caught by a video camera 240, and view data is processed by processor 205, to determine the three-dimensional measurement numerical value of article 290.In some cases, need to use the additional images obtained from other video cameras 240 to determine size.The length of each size is represented as 1/2 inch of (about 12.7 millimeters) increment number of approximate physical length.Such as, the height of 5.5 inches (about 139.7 millimeters) is expressed as numeral 11.15 data bit are used to storage three dimension, and each size uses 5 data bit.Therefore, the maximum length of any single size is 16 inches.This attribute does not have the classification defined in advance, but the physical length of each size is converted into self-defined unit, and resolution is lower than actual weight, and maximal value is 16.Extra figure place can be added process larger maximum length value through confirmation, or increment size value is increased to 1 inch (about 25.4 millimeters) or more, to process larger maximum length value from 1/2 inch (about 12.7 millimeters).

7th physical attribute is " color configuration ".Color configuration attribute is the measurement to eight different color channels, and this passage catches image for what describe article 290.When an electronic image is caught by image capture apparatus (as video camera 240), the numerical data of this image capturing scenes feature by representing forms.The feature of the numerical data caught representated by them forms, and is stored in array of values (also referred to as passage).Graphics standard defines different picture formats, the channel format that each form supports one or more different.Such as, common channel format is " RGB ", represents red, green and blue.This form comprises the Digital Image Data being arranged at least three arrays or passage, the corresponding three kinds of colors of every bar passage: red, green and blue.

In the present embodiment, color configuration attribute is the self-defined channel format based on use eight Color Channels.Article eight, passage is named as: dead color, light color, redness, yellow, green, cyan, blueness and carmetta.The image of scanner 120 by using a video camera 240 to catch article 290, to determine the color configuration of article 290.This view data is processed by processor 205, with the profile of article in recognition image 290.The all view data being positioned at a part of image outside profile are dropped, and remaining like this view data only represents article 290.Then remaining image data format is converted to self-defining eight kinds of channel image forms by processor 205.The view data of every bar passage is normalized to the scope of 0 to 10, then calculates the mean value of every bar passage.The mean value of every bar passage and predetermined threshold value are contrasted.1 attribute being designated as every bar passage, now the mean value of this passage is equal to or greater than threshold value.0 attribute being designated as every bar passage, now the mean value of this passage is less than threshold value.Single data bit needs the attribute storing every bar passage, this means that needs eight data bit store eight passages forming color configuration attribute.

In some embodiments, the color configuration attribute of article 290 comprises the mask of 8 Color Channels.Mask comprises 1 position of eight Color Channels, and mask is used for determining whether corresponding Color Channel will be used for comparison procedure or asterisk wildcard.Such as, masked bits 0 represents and in comparison procedure, does not use corresponding Color Channel numerical value, and masked bits 1 represents employ corresponding Color Channel numerical value in comparison procedure.

It should be noted that the details of above-mentioned attribute can be changed or modified, and still within the scope of the present invention.Such as, one or more elements of color configuration can be changed, or the classification of an attribute can be changed, adds or delete, or the restriction of an attribute can be extended or reduce.In some embodiments, when weight gain is increased to 1 ounce, maximum weight is just increased to 64 pounds from 32 pounds.Some changes may increase the quantity of option, therefore can increase the figure place needing to store possibility option, but can adjust the present invention to adapt to these changes.

In the present embodiment, 40 data bit are needed to store the information of seven physical attributes.After seven the independent physical attributes determining article 290, processor 205 presses predetermined sequence integration seven physical attributes, with the synthesized attribute data record of generate item 290.Synthesized attribute data record is used to identify objects 290 subsequently.This process starts from the field of a known contents mark.This field suitably can comprise nearly 100,000 possible known candidate item.Synthesized attribute data record is used to the decreased number of candidate item to arrive a few.Use the candidate item number reduced to carry out extra process, need the final mark reaching article 290.

Database 225 is stored in the storer 215 of scanner 120.Database 225 comprises multiple data record, and each data record comprises the synthesized attribute data record of a known contents and describes the extraneous information of this known contents, or describes the link of extraneous information of this known contents.Use above-mentioned identical process and seven physical attributes, for this known contents generates synthesized attribute data record.Most of article have multiple side or method to check.Therefore, need to carry out independently record to each major opposing side or view.Use and distribute to the side of this record or view is each record generation synthesized attribute data records.Such as, a box oatmeal has two major opposing sides.They are front and backs of box.All occupy an leading position in these sides, because the area of these two sides wants large more than other lateralareas of box.The record in box front comprises the synthesized attribute data record in front, and the record at the box back side comprises the synthesized attribute data record at the box of cereals back side.Each record also comprises the link of the additional information identifying known contents or the information identifying known contents.The link of use information reduces the size of record.Information about known contents preferably includes an identification code, the Universial Product Code (UPC) of such as identifiable design relative article.This record is stored in database 225.

The scope of the known contents that can be identified by scanner 120 can comprise 100, the article of more than 000.The present invention adopts the synthesized attribute data record of unknown article, and object is that the scope of known contents is reduced to minority known contents, to match with the attribute of unknown article.Once the scope of known contents is down to the candidate item of a group less, this group candidate item only can be used to perform extra process, thus determine the actual characteristic of unknown article.This process can suitably be performed by processor 205.Extra process is beyond scope of the present invention.In some cases, the known candidate item of this group only has one, therefore just can identify unknown article without the need to extra process.But, still suitably can perform extra process and confirm characteristic.

In some embodiments, employ the 8th physical attribute, this attribute is the measurement (if these article are liquid) of the liquid water content of article 290.Radio frequency (RF) energy can be sent to article 290 and measure reflected signal by sending and receiving sensor, thus determines the liquid water content in article 290.This attribute has two classes, i.e. on-liquid or liquid, and the radio-frequency (RF) energy of reflection is used to determine the classification defining article 290.

Refer to Fig. 4 now, this is a process flow diagram, represents that scanner 120 identifies the How It Works of certain article.In the method 400, article 290 are placed on item transport 250, and by the entrance 281 of shell 280 through shell 280, and exposed from shell 280 by outlet 282.Item transport 250 can stop and starting, and makes the one or more sensors 270 in shell 280 have the more time to catch the data of related articles 290 even conversely.Although item transport 250 is depicted as by shell 280 shipping goods 290 on the straight path, other non-directional routes are also predictable.A function of item transport 250 is that the article entering shell 280 are organized into a line article, without the article be arranged side by side in this row, so just can prevent one or more sensor 270 from catching the accurate data relevant to Individual Items.

When article 290 are by shell 280, multiple sensors 270 record the raw data of related articles 290.Subsequently raw data is processed, to generate the value of the multiple physical attributes listed in Fig. 3.In some embodiments, the raw data that process is recorded by one or more sensor 270, to generate a value.Such as, platform scale 235 will deduct tare weight from measurement weight, and the result of generation is the weight of article on scale.But, even if raw data is processed by sensor, but still need extra process to generate the value of multiple physical attribute 305.Raw data or process data (if can obtain from each sensor) are sent to processor 205 and are further processed, and then generate the value of multiple physical attribute 305.

In step 405, article 290 are placed on item transport 250.Article 290 are the parts of the purchase-transaction performed by point of sales terminal 110, and article 290 just can be added in purchase-transaction after needing to be identified.In step 410, item transport 250 sends article 290 to shell 280 by entrance 281.

In step 415, article 290 are transported on platform scale 235.Then platform scale 235 is determined the weight of article 290 and this weight is delivered to processor 205.In some embodiments, item transport 235 comprises three independent sectors, and wherein each part all gives independent control and isolating with other some mechanical, but can be configured article 290 are moved between three parts.The center section of this item transport 250 is installed on the weighing-appliance of platform scale 235.In this configuration, platform scale 235 is weighed to the center section of item transport 250 and article 290.But the weight of item transport 250 center section is known, then deducts from general assembly (TW), to determine the weight of article 290.Once article 290 are positioned on platform scale 235, in order to reduce swing, processor 205 makes the center section of item transport 250 stop mobile article 290.

At step 420 which, each video camera 240 catches the image of article 290.In the present embodiment, when image is captured, article 290 are still positioned on this platform scale 235, do not move.In some embodiments, multiple light source towards the space be positioned on platform scale 235, to illuminate article 290.Light source is controlled by processor 205, and the light that these light sources send can be independently adjustable as closing and sending the various levels such as maximum light.The image captured is transferred to processor 205 and carries out extra process, to determine the physical property values relevant with article 290.Video camera 240 is the equipment aiming at seizure image and design, and therefore may not have all characteristics and the function of consumer video camera.

In step 425, the data of remaining in multiple sensor 270 sensor record related articles 290.This comprises the metal detector 255 that record shows the data containing metal in article 290.Temperature sensor 260 also records the measurement of the temperature to article 290.Ultrasonic transmitter-receiver 245 sends ultrasound wave and enters article 290, and the sound wave of record reflection, after extra process is carried out to this sound wave, will the hardness of article 290 be determined.All record data are all transferred to processor 205 and carry out extra process, to determine the physical property values relevant with each sensor.

In step 430, each sensor in multiple sensor 270 all have recorded the data relevant to article 290, and these data is transferred to processor 205 and carry out extra process.The data that processor 205 uses one or more sensor 270 to record, generate each physical property values of the multiple physical attributes 305 shown in Fig. 3.The value generated meets shown in Fig. 3 measures 310 requirements.Following physical attribute has the value selected from the list of predetermine class: metal, shape, hardness, temperature and color configuration.For all physical attributes, processor 205 processes the data received from the sensor relevant with physical attribute, thus determines primitive attribute, from predetermine class, then select the classification defined with regard to primitive attribute.Each predetermine class may comprise a series of primitive attribute.

Such as, temperature physical attribute has the predetermine class of " freezing ", and this classification is defined as comprising all temperature being equal to or less than 30 °F recorded, and primitive attribute is the temperature of the article 290 recorded by sensor 260.Processor 205 for all be equal to or less than 30 °F record thermal creep stress " freezing " classification.Because temperature physical attribute has 4 predetermine class, and contain all possible temperature, therefore all temperature that records will belong to one of them predetermine class category.Use predetermine class to reduce and store about the quantity of data bit needed for the information of physical attribute, the information in the most relevant information being stored into and identifying needed for certain article can also be limited.Other have the physical attribute of predetermine class to operate in a similar fashion.

Weight and size physical attribute do not have predetermine class, but they need extra process, so that their raw data is converted to customization data, are illustrated in fig. 3 shown below and the above.

In step 435, the synthesized attribute data record of processor 205 generate item 290.In the present embodiment, synthesized attribute data record has the length of 40 data bit, and this data bit comprises the value of all seven physical attributes 305.The position of all values in data record of seven physical attributes 305 is predetermined.In the present embodiment, the all values of seven physical attributes 305 presents with the order identical with the attribute listed in Fig. 3, namely a data bit of metal physics attribute is presented at first in synthesized attribute data record, and eight data bit of color configuration physical attribute are presented at finally in synthesized attribute data record.

In step 440, processor 205 Query Database 225, to select a set of known contents data record.This database 225 comprises multiple known contents data record, and wherein each known contents data record comprises the information defining a known contents, and this information comprises the synthesized attribute data record generated with regard to this known contents.Database 225 can suitably comprise more than 100,000 known contents data record.Utilize the synthesized attribute data record of article 290, processor 205 Query Database 225, then database 225 beams back one group of all known contents data record, and wherein the synthesized attribute data record of known contents conforms to the synthesized attribute data record of article 290.This group data record beamed back generally includes multiple known contents data record.This is that design makes so.A feature of the present invention fast and effeciently a large amount of of article 290 characteristic can may be reduced to one group of very little characteristic.Then the possible characteristic of different skills to this small group is used to be further processed, to determine the characteristic of an energy identify objects 290.

In the present embodiment, database 225 is stored in the storer of scanner 120.This is possible, because the storage of synthesized attribute data record is very efficient, only needs 40 data bit to store.Undersized synthesized attribute data record allows to comprise 100, and the data within Database Requirements 10 megabyte of 000 known contents data record store.But in other embodiments, this database is stored in point of sales terminal 110 or is stored in storage server computer 160.In these embodiments, the synthesized attribute data record of article 290 is sent to point of sales terminal 110 and identifies by scanner 120.If point of sales terminal 110 has database, it performs the inquiry to database, and result is beamed back scanner 120.If storage server computer 160 has database, point of sales terminal 110 sends synthesized attribute data and is recorded to storage server computer 160, then performs inquiry and result is beamed back scanner 120.

Be explained with reference to some preferred embodiments although the present invention is concrete, revised and change in the spirit and scope that the present invention can require at following patent.Such as, when describing particular sensor, may according to particular case, or when new induction technology possesses cost benefit, correspondingly dispose less or more sensor.In addition, when describing example categories in attribute, confirmation may be disposed less or more multi-class.

Claims (17)

1. a computer manner of execution, utilize the physical attribute identification of article for the article of item scanner identification, the method comprises:

The sensor of one or more item scanner is used to calculate the value of each physical attribute of the multiple physical attributes relevant to article;

From multiple physical attribute, identify the physical attribute having predefine classification, wherein all predefine classifications are all based on the preset range of physical property values, and identified for all the predefine classification that physical attribute determines to comprise calculated value;

Generate the synthesized attribute data record comprising each physical attribute of multiple physical attribute;

Multiple synthesized attribute data records that stored of the synthesized attribute data record generated and known contents are compared; And

Choose one group of known contents data record stored, in this record, the synthesized attribute data record that stores of known contents data record selected by each conforms to the synthesized attribute data record generated.

2. the method in claim 1, wherein eachly store the part that synthesized attribute data record is known contents data record, and stored the physical attribute that synthesized attribute data record also represents known contents, and each data record that stored also comprises information for identifying known contents.

3. the method in claim 1 or 2, wherein multiple physical attribute comprises the metal properties shown containing metal in article.

4. the method for any aforementioned claim, wherein multiple physical attribute comprises the temperature property showing article temperature.

5. the method for any aforementioned claim, wherein multiple physical attribute comprises the hardness property showing article rigidity.

6. the method for any aforementioned claim, wherein multiple physical attribute comprises the color configuration attribute showing at least eight kinds of Color Channels.

7. the method for any aforementioned claim, the length of the synthesized attribute data record wherein generated is equal to or less than 8 data bytes.

8. item scanner comprises:

Shell;

The item transport of article can be transmitted via shell; And

Be positioned at the non-contact temperature sensor of shell, this sensor can measure article by temperature during shell.

9. item scanner according to claim 8 also comprises a metal detection sensor, for whether containing metal in detecting article.

10. item scanner according to claim 8 or claim 9, its shell at least comprises a tower, and each tower comprises multiple video camera, these cameras capture to image can calculate shape attribute, size attribute and color configuration attribute through process.

11. item scanner according to any claim 8 to 10 also comprise the platform scale that is positioned at shell.

12. item scanner according to any claim 8 to 11 also comprise a ultrasonic transmitter-receiver, for the hardness property of detecting article.

13. item scanner according to any claim 8 to 12 also comprise a programmed processor, and the information that this processor uses one or more sensor to catch calculates each property value of the multiple physical attributes relevant to article.

14. item scanner according to any claim 8 to 13, its shell defines the entrance receiving article and the outlet withdrawing from article, and all the sensors is all positioned at shell.

15. item scanner according to any claim 8 to 14 also comprise the database that comprises the attribute data relevant with known contents, for the attribute data identify objects helping item scanner to catch according to sensor.

16. 1 kinds of computer manners of execution, comprise the combination in any of any one technical characteristic in claim 1-7 or technical characteristic.

17. 1 kinds of item scanner, comprise the combination in any of any one technical characteristic in claim 8-15 or technical characteristic.