TWI876869B - A trading system that improves the accuracy of matching deductions in unmanned areas - Google Patents

Abstract

The present invention is a transaction system for improving the accuracy of matching deductions in unmanned venues, comprising: a plurality of commodity supply devices, which are arranged on one side of a plurality of directional channels, wherein the channels are defined to form a plurality of shopping blocks corresponding to the positions of the commodity supply devices, so that a plurality of users can move to the shopping blocks and conduct commodity shopping consumption behaviors for the adjacent commodity supply devices; at least one camera device, which records images of the users and generates identification data, wherein the users have an external The appearance recognition feature is used as identity recognition; an identification host is used to receive the identification data and analyze the appearance recognition feature; a usage trajectory database is used to store the historical trajectory information of the user subjects according to the appearance recognition feature; a financial member database is used to store at least one financial transaction tool bound to each of the user subjects; and a point of sale information management device is used to trigger a purchase signal when the user subjects perform consumption behavior, and to judge the identities of the user subjects and the financial transaction tools and execute deductions accordingly.

Description

A trading system that improves the accuracy of matching deductions in unmanned areas

The present invention relates to the relevant technical field of self-service consumption places, and in particular to a transaction system for improving the accuracy of deduction matching in unmanned places, so that when multiple users move on multiple directional channels for consumption, the deduction matching objects can be effectively determined, and possible consumer objects can be appropriately supplemented and consumer objects that should not be consumed can be corrected to improve the accuracy of deduction matching.

With the improvement of technology and the consideration of manpower demand, the current physical consumption business model has derived a variety of consumption models that are not equipped with or equipped with very few service personnel. In order to enable customers to complete overall self-service consumption behavior under this consumption form, it is technically important to efficiently and correctly match the automatic deduction object. In the early days, such consumption places without service personnel were mostly implemented through barcode scanning technology, mainly by consumers manually scanning the product barcode with the automatic deduction system to complete the transaction. Subsequently, with the continuous development of electronic payment technology, such as the rise of RFID (radio frequency identification) and NFC (near field communication) technology, consumers only need to bring their smartphones or payment cards close to the sensor to complete the payment quickly and conveniently. In recent years, artificial intelligence (AI) technologies such as image recognition and machine learning have begun to be applied to the payment judgment process to improve payment security and accuracy.

However, consumption behavior patterns are not always single, and unmanned consumption venues may have a variety of different environmental equipment configurations based on product attributes and lifestyle patterns. For example, for closed unmanned stores, since such stores usually have more product types and consumers have a high degree of freedom in their movement directions without being restricted, it is usually more appropriate to control entry and exit. Therefore, the current practice is to install monitoring devices or identity input devices at the entrance and exit gates to determine the identity of consumers entering and leaving, and to provide an automatic deduction mechanism for subsequent consumption behavior. Since the identity recognition technology of such unmanned stores is relatively simple, the current technical focus of unmanned stores is on how to determine or simplify the relevant judgment mechanism of whether each product has been purchased by consumers.

As time goes by, people's demand for a balance between consumption and time has different changing trends. For example, people usually do not expect to spend too much time shopping for daily necessities. Intuitively, people may prefer to prioritize stores that are fast and have simpler product options for necessary and routine replenishment. However, it is obvious that the aforementioned unmanned store approach will not be effectively applicable to other open consumption venues. For example, in the past, drive-throughs, which enabled consumers to drive through their cars, were a convenient mechanism that greatly improved shopping efficiency. However, if the environment is to be improved to have multiple lanes and multiple product supply nodes in order to maximize supply capacity, the aforementioned unmanned stores that implement detection at a single entrance and exit to perform automatic payment settings would not be easily adaptable.

In view of this, the invention team considered the diversity of future consumption lifestyles, and thus improved upon the past drive-thru consumption patterns, and proposed a new type of self-service consumption venue and model transaction system that does not use entry and exit gates as a control and monitoring method. Specifically, fixed-point monitoring devices can be used in collaboration to identify consumers or driving appearances, and then the consumer's past usage history and the introduction of multiple error-correcting mechanisms can be used to effectively implement matching and supplementation, thereby enhancing the accuracy of payment matching and being used in closed or open unmanned venues. At the same time, the demand for quick purchase of such products with the attribute of grab and go in physical stores is met, greatly improving the quality of the self-service consumption experience.

The main purpose of the present invention is to provide a transaction system that can be used in both open and closed unattended commodity supply places. In this way, no matter the person moves by himself or by driving, the accuracy of matching and deduction can be effectively improved after consumption.

To achieve the above-mentioned purpose, the present invention discloses a transaction system for improving the accuracy of matching deductions in unmanned venues, for multiple users to perform consumption behaviors, including: multiple commodity supply devices, which are arranged on the side of multiple directional channels; wherein the channels are defined to form multiple shopping blocks corresponding to the positions of the commodity supply devices, so that the users can move to the shopping blocks and perform multiple commodity selection and consumption for the adjacent commodity supply devices. behavior; at least one camera device having a monitoring range, the monitoring range covering the selected purchase areas, and recording images of the user subjects to generate identification data; wherein the user subjects respectively have an appearance identification feature for identity identification; an identification host, which is connected to the camera device by telecommunication, for receiving the identification data and further analyzing the appearance identification feature included in the identification data; a usage track database, which is connected to the identification host by telecommunication, and the usage track database The database stores the historical track information of each of the user subjects according to the appearance identification feature, and is used to match and retrieve the historical track information after receiving the appearance identification feature sent by the identification host and return it; a financial member database is connected to the usage track database by telecommunication, and the financial member database stores at least one financial transaction tool bound to each of the user subjects according to the appearance identification feature and the historical track information, and is used to match and retrieve the historical track information according to the appearance identification feature. and a point-of-sale information management device, which is telecommunication-connected to the identification host and the commodity supply devices, and triggers a purchase signal after the user selects one of the commodity supply devices and performs a consumption behavior. The point-of-sale information management device then transmits the purchase signal to the identification host to determine the identity of the user and the financial transaction instrument corresponding to the user, and then transmits it back to the point-of-sale information management device to execute the deduction.

Preferably, after receiving the purchase signal, the identification host performs identity identification and access based on the user subjects on the purchase blocks corresponding to the commodity supply devices; wherein, when the purchase blocks corresponding to the commodity supply devices have no identity identification data and can be accessed, the appearance identification features of the other purchase blocks adjacent to the purchase block without identity identification data are used to identify their identities and access them.

Preferably, when the commodity supply devices are arranged front to back and are arranged on one side, when the purchase area corresponding to the commodity supply devices has no identity identification data and can be accessed, the purchase area in front of or behind the purchase area without identity identification data is identified by the appearance identification features and can be accessed.

Preferably, when the commodity supply devices provide the commodities on both the left and right sides at the same time, one commodity supply device is sandwiched between two shopping blocks; when the shopping blocks corresponding to the commodity supply devices have no identity identification data and can be accessed, the appearance identification features are used to identify the identity of the shopping blocks on the left or right side of the shopping blocks without identity identification data and then access them.

Preferably, the identification host further includes a pre-debugging unit for executing an operation of clearing the appearance identification feature record temporarily stored in the earlier selection block in timing when any of the users identifies the same appearance identification feature in different selection blocks in the earlier timing.

Preferably, the identification host further includes a post-debugging unit for executing an operation of marking the appearance identification feature of any of the users after consumption, and the appearance identification feature of the user is not used within a marking time in the same purchase block.

Preferably, the appearance identification feature of the user is marked, and within a tolerance time in the same purchase block, the appearance identification feature of the user can be used as an exception, and the tolerance time is less than the marking time.

Preferably, the identification host further includes a post-debugging unit for executing an operation of marking the appearance identification feature of any of the users after consumption, and not using the appearance identification feature of the user within a marking time of all the purchase blocks.

Preferably, a plurality of the user entities are grouped together to share the same financial transaction tool to perform deductions.

Preferably, the point-of-sale information management device implements purchase restrictions on the commodities based on the returned appearance identification features and the historical track information.

In summary, the transaction system of the present invention for improving the accuracy of matching payment in unmanned venues is configured to use a multi-channel with directional indication to indicate the direction and route of the user to move to each commodity supply device for consumption, analyze the identity of the user through a camera and a recognition host, and use the various information accumulations of the track database and the financial member database to extract feedback based on the judgment of whether the user is a new or old customer. Finally, the aforementioned information is used in the sales point information management device to match the self-service payment to complete the consumption. In addition, although the monitoring range of the camera device covers all shopping areas, the user may have special circumstances such as bad angles during the movement process, and the appearance visual features may be blocked in some positions, resulting in the situation that they cannot be identified. Therefore, the present invention introduces a variety of error correction mechanisms to effectively implement matching and supplement, thereby enhancing the accuracy of payment matching and being used in closed or open unmanned places. At the same time, it meets the demand for quick purchase of such products with the attribute of pick-up and use in physical stores, greatly improving the quality of self-service consumption experience. In addition, the technical means by which multiple users can use the same deduction tool can further enhance the concept of convenient consumption of corporate or family shared funds. On the other hand, for situations where identity conditions do not match the attributes of the product or where restrictions should be imposed by law, purchase restrictions can be implemented through appearance recognition features and historical trajectory information to reduce property or personal harm derived from improper purchases of goods.

In order to enable people with general knowledge in this field to clearly understand the content of this work, please refer to the following explanation with diagrams.

Please refer to Figures 1 and 2, which are respectively a system block diagram of a preferred embodiment of the present invention; and a schematic diagram of the configuration of a channel and a commodity supply device. As shown in the figure, the present invention proposes a new form of transaction system that improves the accuracy of matching deductions in unmanned venues, so that after a plurality of user subjects 9 perform consumption behaviors, they can effectively match deduction objects, facilitating the overall consumption experience. The transaction system includes: a plurality of commodity supply devices 1, at least one camera device 2, an identification host 3, a usage track database 4, a financial member database 5, and a point of sale information management device 6. The commodity supply devices 1 are arranged on the side of a plurality of channels 8 with directionality. The so-called directionality, as the name implies, means that a certain moving direction must be followed and there is no behavior mode of returning to the original place. The method can also be implemented with the assistance of actual space configuration. For example, the width of the channels 8 can be controlled by the driving moving environment. After the user 9 moves and positions, it is naturally difficult to imagine the concept of moving back and forth again. The channels 8 can be set in closed or open places for implementation. Furthermore, the channels 8 are further defined to form a plurality of shopping blocks 80 corresponding to the positions of the commodity supply devices 7. In other words, the shopping blocks 80 are partial areas within the channels 8 and are respectively arranged adjacent to one side of the commodity supply devices 1, so that the user subjects 9 move to the shopping blocks 80 and purchase multiple commodities 7 from the adjacent commodity supply devices 1. The commodity supply device 1 does not limit the supply of the commodities 7 of any nature, for example, all kinds of daily necessities or food supplies for living things such as humans, pets, and plants, and other things directly supplied in the physical environment for non-living things such as energy replenishment and replacement of various transportation vehicles, etc.; and the commodity supply device 1 is not limited to be implemented by mechanical operation or electronic touch operation. Obviously different from unmanned stores, this type of movement is limited to a certain movement direction and does not allow the user subjects 9 to move back and forth at will. In order to improve sales efficiency, there may be multiple such commodity supply devices 1 that sell the same content of the commodities 7 for the user subjects 9 to consume synchronously. Accordingly, the user subjects 9 can choose the commodity supply device 1 to be used in the indicated movement direction for operation, so logically there is no need to move back and forth, because any operating position can be used by the user subjects 9 to obtain the same commodities 7. In addition, the meaning of the user subjects 9 can be simply a natural person object or a driving moving object, or any object with the appearance identification feature 90, and is not limited to the above examples.

Next, the camera device 2 has a monitoring range that covers the selection areas 80, and records images of the user subjects 9 to generate identification data, wherein the user subjects 9 each have an appearance identification feature 90 for identity identification, and the identification host 3 is telecommunication-connected to the camera device 2 to receive the identification data and further analyze the appearance identification feature 90 included in the identification data. Specifically, the camera device 2 can utilize a general fixed camera, and when the subject 9 such as a natural person or a vehicle moves to the purchase area 80, the camera device 2 will capture and record the facial features of the natural person as much as possible through the monitoring range; or the license plate features of the vehicle as the appearance recognition feature 90 to determine its identity.

Furthermore, the usage trajectory database 4 is telecommunication-connected to the identification host 3, and the usage trajectory database 4 stores and corresponds to one historical trajectory information 40 of each of the user subjects 9 according to the appearance identification feature 90, and is used to match and retrieve the historical trajectory information 40 and return it after receiving the appearance identification feature 90 transmitted by the identification host 3; the financial member database 5 is telecommunication-connected to the usage trajectory database 4, and the financial member database 5 stores and corresponds to at least one financial transaction tool 50 bound to each of the user subjects 9 according to the appearance identification feature 90 and the historical trajectory information 40, and is used to match and retrieve according to the appearance identification feature 90 and return the financial transaction tool 50. As for the point-of-sale information management device 6, it is connected to the identification host 3 and the commodity supply devices 1 by telecommunications, and a purchase signal is triggered after the user subjects 9 select one of the commodity supply devices 1 to perform consumption behavior. The purchase signal is then transmitted by the point-of-sale information management device 6 to the identification host 3 to determine the identity of the user subjects 9 and the financial transaction instrument 50 corresponding to the user subjects 9, and then sent back to the point-of-sale information management device 6 to execute subsequent payment deductions.

In other words, when the user 9 moves to the purchase area 80 adjacent to the commodity supply device 1, the user 9 selects a certain commodity 7 and triggers the purchase signal. At this time, the sales point information management device 6 informs the identification host 3 that it wants to retrieve the identity of the user 9, for example, through license plate recognition or face recognition as the appearance recognition feature 90, and then cross-checks the historical track information 40 corresponding to the appearance recognition feature 90 and the financial transaction tool 50. The historical track information 40 includes records of past visits, non-visits, and past consumption situations; the financial transaction tool 50 is the payment method bound to the appearance recognition feature 90. In this way, after the user 9 completes the consumption, it can naturally automatically call the matching deduction object according to the aforementioned information to quickly complete the payment. Of course, even if it is the first visit and no financial transaction tool is bound, the usage track database 4 will also record it. Although the automatic deduction mechanism cannot be used at that time, there is still a corresponding record for the user 9 to pre-judge or inform the user that the financial transaction tool 50 can be bound to facilitate consumption the next time. In order to respond to various consumption behavior patterns in reality, for example, the concept of sharing reserve funds or stored-value cards is given to enterprises or families, and the same payment tool can be shared by the members of the enterprise or family to complete the deduction. In other words, a plurality of the user subjects 9 can be set up as a group and share the same financial transaction tool 50 to execute deductions, thereby enhancing the convenience of using the transaction system of the present invention. Incidentally, since the focus of the present invention is on the automatic deduction matching mechanism, rather than how to determine whether the product has been selected, this part can be implemented according to the existing mechanism, such as using weight or occlusion detection, etc., which will not be elaborated here. Moreover, the types of the products 7 are different, and the trigger signal may also be set differently according to the actual properties of the products 7. For example, the product supply device 1 provides pure liquid items and does not provide packaging based on environmental protection considerations. In this case, the user subjects 9 must fill their own containers. During the process, when the user 9 clicks the touch panel of the commodity supply device 1 to select or presses the physical button for providing the commodities 7, it can be used as the activation judgment point of the trigger signal.

Furthermore, the point-of-sale information management device 6 can implement purchase restrictions on the commodities 7 according to the appearance identification features 90 and the historical track information 40 sent back. For example, the user 9 may always choose American black coffee in the past, and the possible reason is simply due to simple preference. At this time, if the user mistakenly chooses a drink with milk, the historical track information 40 will prompt the user 9 to avoid wrong consumption. Of course, for more serious prohibition situations, such as providing the contents of the commodities 7 will cause irreversible personal or property damage, the system can also strictly implement purchase restrictions according to the historical track information 40 of the user 9, that is, the machine will not provide the physical supply of the commodities 7.

Please refer to Figure 2 again. As mentioned above, since the camera device 2 may be blocked or misoperated during image recording, the point-of-sale information management device 6 cannot find the corresponding user subject 9 in the corresponding purchase area 80 when the user subject 9 performs consumption behavior in one of the commodity supply devices 1 to trigger the purchase signal. At this time, a matching and supplementing mechanism should be implemented. In short, after receiving the purchase signal, the identification host 3 performs identity identification and access according to the user subjects 9 on the purchase blocks 80 corresponding to the commodity supply devices 1; however, when the purchase blocks 80 corresponding to the commodity supply devices 1 have no identity identification data and can be accessed, the appearance identification features 90 are used to identify the identities of other purchase blocks 80 adjacent to the purchase block 80 without identity identification data and access them. Furthermore, in order to improve the accuracy of the missing matching, different missing judgment mechanisms can be formed according to the arrangement of the commodity supply devices 1 and the spatial placement of the commodities 7. Among them, when the commodity supply devices 1 are arranged front to back and are configured to supply the commodities 7 on one side, when the purchase area 80 corresponding to the commodity supply devices 1 has no identity identification data and can be accessed, the appearance identification feature 90 is used to identify the identity of the purchase area 80 in front of or behind the purchase area 80 without identity identification data and access it. There are many reasons for this situation to happen in experience. For example, it is common to operate the equipment at the nearest location, but there is no definite standard for the so-called nearest location. Moreover, the purchase area 80 is a virtual frame area set by the system, so in reality, consumers have no awareness of the purchase area 80. Naturally, during consumption behavior, they may stay at the commodity supply device 1 and the purchase area 80 that do not correspond to each other. At this time, the closest mechanism is used to fill in the actual consumer position to improve the accuracy of matching deduction.

In order to strengthen the supply capacity of the commodities 7 in the overall field, the commodity supply devices 1 can also provide the commodities 7 on both the left and right sides at the same time, and one commodity supply device 1 is sandwiched between two purchasing blocks 80; and when the purchasing block 80 corresponding to the commodity supply devices 1 has no identity identification data and can be accessed, the purchasing block 80 on the left or right side of the purchasing block 80 without identity identification data is identified by the appearance identification feature 90 for identification and access. Specifically, in order to reduce the error in replenishment judgment, when the commodity supply device 1 is configured to supply the commodities 7 on both sides at the same time, the two commodity supply devices 1 are not arranged in a close arrangement front and back, so as to avoid that there are multiple alternative appearance visual features 90 available for selection when the system makes replenishment judgment, thereby causing errors in deduction matching.

Please refer to Figure 3, which is a system block diagram of another preferred embodiment of the present invention. As described above, in addition to the matching mechanism for supplementing the missing parts, the present invention can also improve the matching accuracy by introducing a debugging mechanism as appropriate. As can be seen from the figure, in the debugging mechanism at the pre-consumption stage, the identification host 3 further includes a pre-debugging unit 30 for executing the operation of clearing the appearance recognition feature 90 record temporarily stored in the earlier selection block 80 when any of the user subjects 9 recognizes the same appearance recognition feature 90 in the different selection blocks 80 in the previous and subsequent timings. That is, the user subjects 9 may move to different sides of the commodity supply device 1, but the contents recorded in each selection block 80 during the movement process will be judged with priority given to the last stop position. On the other hand, in the debugging mechanism of the post-consumption stage, the identification host 3 further includes a post-debugging unit 32 for executing the operation of marking the appearance identification feature 90 of the user subject 9 after any of the user subjects 9 consumes, and the appearance identification feature 90 of the user subject 9 is not used within a marking time in the same selection block 80. Among them, the marking time refers to the pre-set time setting when the consumer completes the consumption of the commodity 7 and leaves under normal circumstances. For example, after a user 9 completes the purchase of a coffee drink in a container brought by the user in the shopping area 80 numbered A, theoretically, the shopping area 80 numbered A should not be considered as a deduction target within 2 minutes after the consumption. However, this situation is usually not in line with normal consumption and should be excluded. However, considering the short-term repeated consumption behavior of some users 9 in the same shopping area 80, a tolerance time is introduced at this time to temporarily unmark the user 9 that is marked as not being a deduction matching target. That is, when the appearance identification feature 90 of the user 9 is marked, the appearance identification feature 90 of the user 9 can be used as an exception within the error tolerance time of the same purchase block 80, and the error tolerance time is less than the marking time, for example, the error tolerance time is 20 seconds and the marking time is 120 seconds. In order to enhance the probability of error correction in the post-consumption stage, the post-error correction unit 32 included in the identification host 3 can also be used to perform the operation of marking the appearance identification feature 90 of the user 9 after any of the user 9 consumes, and within a marking time of all the purchase blocks 80, the appearance identification feature 90 of the user 9 will not be used. In other words, the difference between the two debugging mechanisms at the post-consumption stage mentioned above mainly lies in the debugging mechanism that prevents the same purchase block 80 or all the purchase blocks 80 from being repeatedly identified as objects to be deducted within the empirical marking time after consumption.

In summary, the transaction system of the present invention for improving the accuracy of matching payment in unmanned venues is configured to use a multi-channel with directional indication to indicate the direction and route of the user to move to each commodity supply device for consumption, analyze the identity of the user through a camera and a recognition host, and use the various information accumulations of the track database and the financial member database to extract feedback based on the judgment of whether the user is a new or old customer. Finally, the aforementioned information is used in the sales point information management device to match the self-service payment to complete the consumption. In addition, although the monitoring range of the camera device covers all shopping areas, the user may have special circumstances such as bad angles during the movement process, and the appearance visual features may be blocked or misjudged in some positions, resulting in inability to identify or errors. Therefore, the present invention introduces a variety of error correction mechanisms to effectively implement matching and supplement, thereby enhancing the accuracy of payment matching and being used in closed or open unmanned places. At the same time, it meets the demand for quick purchase of such products with the attribute of pick-up and use in physical stores, greatly improving the quality of self-service consumption experience. In addition, the technical means by which multiple users can use the same deduction tool can further enhance the concept of convenient consumption of corporate or family shared funds. On the other hand, if the identity conditions do not match the attributes of the product or the law should restrict it, purchase restrictions can be implemented through appearance recognition features and historical trajectory information to reduce property or personal hazards derived from improper purchases of goods.

The above is only an illustration of the preferred embodiment within the scope of the patent application of the present invention, and the scope of rights of the present invention cannot be limited by the content of this embodiment; therefore, any changes or modifications made within the scope of equality of the present invention should still be included in the scope of the patent application of the present invention.

1: Product supply device 2: Camera device 3: Identification host 30: Front-end error correction unit 32: Rear-end error correction unit 4: User track database 40: Historical track information 5: Financial member database 50: Financial transaction tools 6: Point of sale information management device 7: Product 8: Channel 80: Purchase area 9: User subject 90: Appearance recognition feature

Figure 1 is a system block diagram of a preferred embodiment of the present invention. Figure 2 is a schematic diagram of the channel and product supply device configuration of a preferred embodiment of the present invention. Figure 3 is a system block diagram of another preferred embodiment of the present invention.

1: Product supply device

2: Camera device

3: Identify the host

4: Use the track database

40: Historical track information

5: Financial member database

50: Financial trading tools

6: Point of sale information management device

7: Commodities

Claims (10)

A transaction system for improving the accuracy of matching deductions in unmanned venues, for multiple users to perform consumption behaviors, including: Multiple commodity supply devices, which are arranged on the side of multiple directional channels; wherein the channels are defined to form multiple shopping blocks corresponding to the positions of the commodity supply devices, so that the users can move to the shopping blocks and perform multiple commodity selection consumption behaviors for the adjacent commodity supply devices; At least one camera device has a monitoring range, which covers the selected areas, and records images of the users to generate identification data; wherein the users each have an appearance identification feature for identity identification; An identification host, which is connected to the camera device by telecommunication, is used to receive the identification data and further analyze the appearance identification feature included in the identification data; A usage track database, which is connected to the identification host by telecommunications. The usage track database stores the historical track information of each of the user subjects according to the appearance identification feature, and is used to match and retrieve the historical track information after receiving the appearance identification feature sent by the identification host and return it; A financial member database, which is connected to the usage track database by telecommunications. The financial member database stores at least one financial transaction tool bound to each of the user subjects according to the appearance identification feature and the historical track information, and is used to match and retrieve the financial transaction tool according to the appearance identification feature and return it; and A point-of-sale information management device is telecommunication-connected to the identification host and the commodity supply devices, and triggers a purchase signal after the user selects one of the commodity supply devices and performs a consumption behavior. The purchase signal is then transmitted by the point-of-sale information management device to the identification host to determine the identity of the user and the financial transaction instrument corresponding to the user, and then sent back to the point-of-sale information management device to execute the deduction. A transaction system as described in claim 1, wherein, after receiving the purchase signal, the identification host performs identity identification and access based on the user subjects on the purchase blocks corresponding to the commodity supply devices; wherein, when the purchase blocks corresponding to the commodity supply devices have no identity identification data and can be accessed, the appearance identification features of the other purchase blocks adjacent to the purchase block without identity identification data are used to identify their identities and access them. In the transaction system as described in claim 2, when the commodity supply devices are arranged in a front-to-back manner and are arranged in a unilateral manner, when the purchase area corresponding to the commodity supply devices has no identity identification data and can be accessed, the purchase area in front of or behind the purchase area without identity identification data is identified by the appearance identification features and accessed. A transaction system as described in claim 2, wherein when the commodity supply devices provide the commodities on both the left and right sides, the commodity supply devices are arranged to be sandwiched between two shopping blocks; when the shopping block corresponding to the commodity supply devices has no identity identification data and can be accessed, the identity of the shopping block on the left or right side of the shopping block without identity identification data is identified by the appearance identification features and then accessed. A trading system as described in any one of claim items 1 to 4, wherein the identification host further includes a front-end debugging unit for executing an operation of clearing the appearance identification feature record temporarily stored in the earlier purchase block in the timing when any of the users identifies the same appearance identification feature in different purchase blocks in the previous and subsequent timings. The transaction system as described in claim 5, wherein the identification host further includes a post-debugging unit for executing an operation of marking the appearance identification feature of any of the users after consumption, and the appearance identification feature of the user is not used within a marking time in the same purchase block. A trading system as described in claim 6, wherein the appearance identification feature of the user is marked, and the appearance identification feature of the user can be used as an exception within a tolerance time in the same purchase block, and the tolerance time is less than the marking time. A transaction system as described in claim 5, wherein the identification host further includes a post-debugging unit for executing an operation of marking the appearance identification feature of any of the users after consumption, and the appearance identification feature of the user is not accessed within a marking time of all the purchase blocks. A trading system as described in claim 5, wherein a plurality of such user entities are grouped together to share the same financial transaction instrument to perform deductions. A transaction system as described in claim 5, wherein the point-of-sale information management device implements purchase restrictions on the commodities based on the returned appearance identification features and the historical trajectory information.