A commodity display and transaction method and device based on AI recognition technology

Abstract

The present invention relates to the technical field of e-commerce platforms, and in particular to a commodity display and transaction method and device based on AI recognition technology, comprising: S1, receiving commodity information data collected by a collection device or device through a collection module; S2, performing AI recognition processing on the collected commodity information data through an AI recognition processing module; S3 protecting the privacy of merchants through a privacy protection module; S4, displaying commodities and trading commodities through a transaction support module; it identifies and processes complex commodity information through the collection module, AI recognition processing module, privacy protection module and transaction support module, automatically enters and updates commodity information, and displays the physical information of commodities to consumers after protecting the privacy of merchants, optical indication facilitates commodity positioning and selection, realizes the simultaneous sales of commodities online and offline, reduces the time occupied by merchants without affecting their offline sales, and achieves the purpose of moving the market online.

Description

Technical Field

[0001] The present invention relates to the technical field of e-commerce platforms, and in particular to a commodity display and transaction method and device based on AI recognition technology. Background Art

[0002] In the past, consumers had to go to markets and supermarkets to buy vegetables, fruits, fresh fish and meat, delicacies and other foods in their daily lives. However, now e-commerce platforms have emerged, such as an Internet-based e-commerce platform system disclosed in the invention patent with announcement number CN110414902B and an e-commerce platform based on blockchain technology disclosed in the invention patent with announcement number CN111931212B. They are very mature and online shopping has become very common.

[0003] However, existing e-commerce platforms mainly target products with stable quality, long shelf life, and infrequent price changes, while agricultural and sideline products are of many types, with different quality and large daily price changes. In addition, the pictures and videos of products displayed on existing e-commerce platforms cannot reflect the true status of the current products, and it is impossible to specify the products to be purchased, which makes consumers unable to purchase with confidence. Although live video broadcasting can visually see the products, it takes up a lot of time of the anchor and is prone to privacy leakage. Due to the high threshold of live broadcasting, merchants are also busy with offline sales, and it is difficult to synchronize online and offline, which makes it difficult to sell agricultural and sideline products online. Therefore, a product display and transaction method and device based on AI recognition technology are urgently needed to improve the above problems. Summary of the invention

[0004] In order to solve the above technical problems, the present invention simulates real transaction scenes and provides a commodity display and transaction method and device based on AI recognition technology, which can identify and process complex commodity information, automatically enter and update commodity information, protect the privacy of merchants, and display the physical information of commodities to consumers after protecting the privacy of merchants. Optical indication facilitates commodity positioning and selection, realizes simultaneous online and offline sales of commodities, reduces merchants' time occupation, and achieves the purpose of moving the market online.

[0005] A commodity display and transaction method based on AI recognition technology of the present invention comprises:

[0006] S1. Receive commodity information data collected by a collection device or apparatus through a collection module;

[0007] S2. Perform AI recognition processing on the collected commodity information data through the AI ​​recognition processing module;

[0008] S3, protect the privacy of merchants through the privacy protection module;

[0009] S4. Displaying commodities and trading commodities through the transaction support module;

[0010] Through the collection module, AI recognition and processing module, privacy protection module and transaction support module, complex product information is identified and processed, automatically entered and updated, and the physical information of the product is displayed to consumers. Optical indication facilitates product positioning and selection, and realizes the simultaneous online and offline sales of product information, reducing merchants' time without affecting their offline sales, thereby achieving the goal of moving the market online.

[0011] Preferably, the product information data:

[0012] Receive product appearance, name, price, device posture and position information recorded manually or automatically according to time intervals or scanning areas and paths, for entry and update of product information and indication and viewing of physical products. The collected product information data includes audio, video, image, QR code, name, price information, origin information and other data.

[0013] Preferably, the posture position information:

[0014] According to the situation of on-site product selection, the information of the relative spatial position relationship between the equipment and the current item is collected, including the rotation angle, pitch angle, focal length, number, feature points and other information of the collection equipment; it is used for the collection and update of product information. If the product is out of stock, it will be automatically taken off the shelf. If the product is blocked, it will be re-recorded after a delay. If the equipment is not fixedly installed, gesture guidance can be used in conjunction with voice and optical indication to complete the corresponding product indication and entry tasks; it is used for consumers to purchase special products (divisible, selectable non-mixed products), and optical indication will indicate the product to consumers or merchants based on the posture and position information; when the optical indication is manually entered into the product information, the posture and position information locates the product and guides the merchant to enter the corresponding product information.

[0015] Preferably, the optical indication:

[0016] Optical indication is a light generator (such as a laser transmitter, etc.) emitting visible light for illumination indication, using changing light to represent different information, such as light color, flashing frequency, etc., to assist in product positioning indication, manual product entry guidance, product search, order, browsing, delivery, acquisition equipment standby, startup, equipment acquisition field of view facing position, cruising, anti-theft protection, etc. The sounder can cooperate to emit corresponding prompt sounds;

[0017] When automatically entering products, the optical indicator light indicates the scanning path, area, and scanning status.

[0018] When manually inputting and guiding the entry of product information, optical indicators can be used to illuminate the products to guide merchants to enter the corresponding information in voice, pictures, and text.

[0019] When consumers point to products or purchase special products (divisible, non-mixed, durian, delicacies), consumers drag the indication target on the image, and the optical indication follows the positioning and indication of the product. After the consumer selects the product or draws the product selection area graphic, the selected product information is generated according to the posture position information;

[0020] The merchant searches for and processes the indicated product. 1. The optical indicator indicates the indicated product according to the generated indicated product information. 2. The merchant selects the product according to the graphic. 3. The merchant selects the product according to the product number, so as to meet the merchant's need to quickly search for and process the selected product.

[0021] The optical indicator can follow the merchant to the metering equipment and complete the operations such as uploading the metering information during the metering process.

[0022] Preferably, the scanning path and area:

[0023] Automatic drawing: Scanning paths and areas automatically draw line segments on scene images according to features, and generate path data and scanning area data by combining images and device posture and position information;

[0024] Manual drawing: The product scanning path and area are manually drawn by the merchant on the captured image, and the path data or area data is generated by combining the image and the device posture and position information;

[0025] Manually guided device scanning: used for mobile merchants or groups with unstable merchandise displays and those who are too old to operate. The scanning path for merchandise often changes. Merchants can use the optically indicated light spots in conjunction with the end points of the column (finger or root end point sliding to indicate the collection device) to follow and combine voice and image recognition to generate the guided path and area. When entering merchandise information, merchants complete the collection by entering the corresponding voice, image, text and other information based on the optical indication of the merchandise.

[0026] The scanning path generates corresponding quantities and numbers (not unique) according to the merchant's product display and site conditions. The scanning area is divided into display area and non-display area. The display area is mainly used for product information collection and display, and the non-display area is mainly used for privacy protection.

[0027] Automatic scanning: The device scans and records the collected data according to the set path data, and processes the collected data accordingly according to the area data; searches for the indicated physical goods according to the posture and position information; performs privacy protection, anti-theft protection, business status monitoring, equipment power on and off, etc. according to privacy protection and related settings. Scanning can also be guided manually.

[0028] Preferably, the manually guided scanning:

[0029] Manually guided device scanning: Merchants use the end points of fingers and wood roots to slide with the optical indicator light spots to guide the collection device to follow, and combine voice and image recognition to generate product collection data and scanning paths and scanning areas. When entering product information, merchants enter corresponding voice, image, text and other information according to the instructions of the optical indicator on the product to complete the scanning and collection. The collection device can also be controlled by the pan-tilt system to manually collect products one by one through the collection device, and generate a scanning path based on the posture and position of the product equipment;

[0030] After the automatic or manual collection is completed and a scanning path is generated, the collection device can update the commodity status information regularly according to the scanning interval set by the merchant and the scanning path.

[0031] Preferably, the AI ​​recognition processing method is:

[0032] The commodity information data collected by the collection device is analyzed, identified and matched with the data in the database through AI recognition technology, and the commodity is split or the information is merged to generate commodity display information of a single product. The AI ​​recognition processing of the commodity information collection specifically includes the following:

[0033] Human-object modeling, model training and model learning, feature extraction and recognition of various types of product information data collected through recognition processing, splitting the recognized products, listing the product appearance information in units of single products, identifying the graphic, text and audio data related to the products, identifying the product name, price, origin, production and sales time and other additional product information, and merging the product appearance information with the product additional information to generate the product display information of the single product; the product display information is displayed in the form of images, videos, animations, models, texts, voices, etc., and the display information includes: product appearance, name, price, origin, production date, number, AI salesperson, product layout, store scene, etc.;

[0034] Identification processing content and methods: including person recognition, object recognition, image recognition and segmentation, speech recognition, etc.; by denoising, color conversion, scaling, cropping, rotating, moving, enhancing, sharpening, feature extraction, frame comparison, segmentation, labeling, center point posture extraction and other operations on the collected information, and comparing with the established model library for recognition, classification, evaluation and other processing to complete the recognition, generate model data, text information, QR code information, product information, output comparison results, etc. Models include: person model, product model, environment model, facility model, scene model, voice model, AI salesperson model, etc.; establish and train models by: extracting feature points, shapes, colors, textures, reflections, bumps, scales and other parameters from related materials to create model data, and then compare and train with similar model data in large quantities, optimize model data to improve the accuracy of model data, and store it in the model library; application tools such as: OpenCV library for preprocessing, enhancement, recognition, TensorFlow to build a deep neural network model for recognition, etc.; denoising, enhancing, endpoint detection, training, matching and recognition, Complete speech recognition and result output, such as Google, iFLYTEK and other speech recognition engines;

[0035] Preferably, the privacy protection:

[0036] It is used to modify or block the people, objects or contents of the inconvenient display areas set by the merchant (such as private items and materials in the inner room or on the desk), turn on or off the equipment at a scheduled time, control the power on and off of the equipment with photosensitive sensors (such as turning off the lights and locking the door at night), identify the business status and objects: such as the switch status of the business door or the status of the upper and lower locks to control the power on and off of the equipment, etc.; set the protection area, privacy protection content, people or objects, etc. The area setting is: draw the collection area grid or the non-collection area grid on the picture that records the scanning path; divide the collected data into display areas and non-display areas, do not record or modify the information in the non-display area, distinguish, encode or replace non-trading people or objects in the display area, and protect privacy information. , protection information, and third-party information are not stored or displayed; the equipment is automatically started or shut down according to the business hours or start and close times set by the merchant; the merchant has the highest authority and sets up multi-level network information security precautions. The operation and protection rules of all equipment and devices are authorized and formulated by the merchant. The goods are viewed according to the set browsing time limit, browsing method, queuing rules, and following rules (used for merchants to introduce recommended goods, etc.). Any equipment control instructions not authorized by the merchant will not be executed; relevant rules and parameters can be set and processed through privacy protection modules or external programs (such as device clients, third-party software, etc.). Merchants formulate anti-theft and alarm rules in combination with equipment and devices, and cooperate with identification technology for security detection and protection, focusing on non-business hours.

[0037] Preferably, the method for processing the business status identification object is:

[0038] Business status identification objects include: business doors, locks, etc.

[0039] Used to monitor the business status of merchants, identify whether they are open, and enable or disable privacy protection and anti-theft protection, etc.

[0040] According to the time interval set by the merchant or the change of the light in the place, the information of the doors or locks in the designated area is scanned, and the state of the doors or locks in business and non-business hours is learned and trained to achieve the recognition of the switch state of the doors or locks and execute the corresponding settings; the recognition process is: formulate the associated doors or locks, take pictures of the open and closed states, and perform denoising, color conversion, scaling, cropping, rotation, movement, enhancement, sharpening, feature extraction, frame comparison, segmentation, labeling, extracting the center posture position feature points to establish a model, training the model, recognition, evaluation and other business status identification;

[0041] Application scenarios such as:

[0042] 1. If the rolling door is down, it means the business is closed, the external transmission of the venue information is closed, the relevant privacy protection is activated, and the anti-theft monitoring alarm is activated;

[0043] 2. For locked places, the business status information is output through modeling, learning, training, and identification according to different lock types. When the diagnosis is locked, the external transmission of the place information is closed, and relevant privacy protection and anti-theft monitoring alarm are activated;

[0044] 3. Anti-theft alarm monitoring can be clear monitoring or non-clear monitoring. Non-clear monitoring means that the equipment blurs or codes the transmitted monitoring image information, and the detector (merchant or illegal intruder, etc.) sees the picture as a non-clear image. This measure can further increase the merchant's protection against privacy leakage security. During the anti-theft alarm period, monitoring is carried out according to the anti-theft rules. After an alarm event occurs, the scene information is stored and the alarm information is sent to the merchant.

[0045] Preferably, the method of product splitting and information merging is:

[0046] The complex product information recorded is split into multiple single product information, matched with product-related information and entered into the database to generate detailed single product display information. After review by the merchant, it is used to display and trade products. The merchant reviews and modifies the generated product information; the merchant can also manually collect and enter relevant product information.

[0047] Preferably, the method for indicating goods:

[0048] When consumers point to and select products, the optical indicator uses a light generator to emit visible light to locate and indicate the product according to the posture and position information of the product selected by the buyer. By changing the color or strobing of the light, the merchant is indicated to different states (such as browsing, connecting, placing an order, turning on and off, etc.), and the sounder emits a corresponding prompt sound;

[0049] Consumers draw a graphic on the product image to indicate the area of ​​the selected product or select the desired product based on the product number, and the merchant quickly processes it based on the selected area or number;

[0050] The collection equipment should be installed on the top (vertical to the items), outside the stall facing the stall (facing the merchants and facing away from consumers), etc., to facilitate identification and protection;

[0051] For merchants whose collection equipment is not fixedly installed (such as car owners in wholesale markets and mobile street stalls), corresponding power supplies and voltage adapters are provided. Merchants can use gesture guidance methods combined with voice and optical indications to enter information or indicate products.

[0052] Preferably, the method for viewing the physical product is:

[0053] When the merchant processes an order, the optical indicator light indicates the goods to the merchant according to the device posture information of the selected goods, so that the merchant can quickly find and process the goods;

[0054] Consumers are authorized by merchants to connect to the merchant's collection device within a limited time, instructing to view the actual product and communicate with the merchant or AI salesperson online, specifically:

[0055] Consumers connect to the merchant's collection device through the merchant's authorization or set rules. The collection device quickly indicates the location of the pre-ordered product or area based on the posture and position information of the device where the product is located, and sends back the physical image or video of the product or allows the product to be viewed online for a limited time. The consumer can connect to the merchant or AI salesperson at the same time. The merchant communicates with the consumer through audio and video via voice equipment (headphones, mobile phones, tablets, etc.), and the AI ​​salesperson can automatically answer questions and make sales.

[0056] Preferably, the privacy protection method:

[0057] Control equipment by identifying business status, objects, and timing, and perform privacy information protection, anti-theft alarm, and other processing according to the protection rules for areas, people, and objects set by merchants;

[0058] Preferably, the method for displaying and trading goods:

[0059] Display products: push nearby products or markets based on the product name selected by the consumer and the relative distance between the two parties. The display mode is in the product mode or market mode.

[0060] Trading goods: Merchants set wholesale and retail quantities and prices based on their own business nature, and consumers choose the corresponding wholesale or retail model based on their own characteristics and demand;

[0061] The specific methods for displaying and trading goods are as follows:

[0062] Agricultural and sideline products have four links: production, wholesale, retail and consumption, involving production groups, wholesale groups, retail groups, consumer groups, etc.

[0063] According to the supply and demand relationship of the entire industry chain, the commodity trading model is set as: wholesale and retail model; merchants set the wholesale and retail quantities and corresponding prices according to their own business nature; consumers choose the corresponding purchasing model according to their own characteristics and demand;

[0064] According to the name of the product selected by the consumer and the relative distance between the two parties, nearby products or markets are pushed. The product display and browsing mode is:

[0065] The commodity model is suitable for consumers who have a clear idea of ​​the types and quantities of commodities to be purchased. In this model, the pre-order commodities are listed and pushed to consumers for their selection.

[0066] The market model is suitable for consumers who are not clear about the goods to be purchased or buy goods randomly. In this case, nearby merchants will be displayed to consumers, and consumers will enter the merchant page to purchase goods randomly.

[0067] Preferably, the transaction support module:

[0068] Location delivery unit: pushes goods and matches the best and shortest delivery service in the same area according to the relative positions of the two parties to the transaction, and manages the delivery and receipt of goods, specifically;

[0069] Transaction support unit: records and manages the commodity transaction process;

[0070] Client: facilitates information browsing, publishing and communication between both parties of the transaction;

[0071] Traceability credit unit: Pesticides, additives, origins of the product production process, transaction process related information, merchant transaction integrity, service satisfaction, etc. are recorded in graphic and text traceability information to prepare for the retrieval and merging of relevant information during AI recognition processing;

[0072] Database: record, store and call relevant data of the system.

[0073] A commodity display and transaction device based on AI recognition technology, characterized in that it includes a central processing unit, a memory, an optical sensor component, a communication component, a command control component, a built-in program component and a bus, and the central processing unit, the memory, the optical sensor component, the communication component, the command control component and the built-in program component are all connected to the bus;

[0074] Preferably, the optical sensor assembly includes a passive or active type (such as an image sensor, a three-dimensional laser scanning sensor, a photosensitive sensor, a light generator, etc.);

[0075] Preferably, the communication component:

[0076] Communication components include: network module, Bluetooth module, voice component, etc. Among them, the network and Bluetooth components have the ability to connect to multiple devices such as headphones, electronic scales, label machines, cash registers, etc.; the voice component can combine control commands with other components to allow the device to complete various tasks, such as executing product information entry and update, power on and off, order reports, product positioning, etc.

[0077] The central processing unit controls and manages the memory, optical sensor component, communication component, instruction control component and built-in program component. The memory stores data, the optical sensor component scans and records goods and product information, the communication component transmits data, the instruction control component receives task instructions sent by consumers, performs mechanical movement, optical indication, sound and electrical action, the built-in program component processes data and instructions, and the external program sends commands to the device and sets relevant device parameters.

[0078] Preferably, the optical indication comprises:

[0079] Optical indicators include: light generators (such as laser transmitters) are preferably installed around the optical image model acquisition sensor and integrated in the same direction; used to emit visible light to illuminate the designated items, used to instruct buyers to select or select goods, and used by merchants to find goods based on the goods selected in the order. For example: a constant light indicates that the buyer is selecting a product, and a flashing light indicates that the product has been selected;

[0080] The indicator light emits different colors and flashes to indicate different transaction states, such as standby, power on, viewing products, pending orders, browsing, new orders, and pending shipments.

[0081] Preferably, mechanical movement includes controlling the servo motor to drive the device to rotate, pitch, focus and other actions and recording related parameters to generate attitude position value data.

[0082] Preferably, the built-in program component is provided with a built-in program;

[0083] Built-in programs include:

[0084] Information processing: The program pre-processes the protection area, person, anti-theft, identification and other information in the collected data according to the rules;

[0085] Equipment posture and position processing: extracting the rotation angle, pitch angle, focal length, number, feature points and other information of the acquisition equipment device to determine the relative spatial position relationship between the acquisition equipment device and the current object;

[0086] Task control instruction set: formulate dedicated control commands for interaction between various software platforms. By receiving task instructions, the device is controlled to perform related actions to complete related tasks or feedback related information. The instruction set is compatible with Onvif and other protocols.

[0087] Preferably, the external program includes: a device client and a third-party software platform; the external program sets or modifies device-related parameters, sends instructions to the device to control the equipment to perform related actions to complete related tasks and feed back related information data.

[0088] Compared with the prior art, the present invention has the following beneficial effects:

[0089] 1. The present invention uses AI to identify, split, process, and generate individual product display information for product information of various types, different qualities, and frequent price changes, automatically updates product information or views physical products online, reduces the workload of merchants, reduces the time occupied by merchants, and truly achieves simultaneous offline and online sales; after privacy protection processing, the real status information of the product is presented to consumers, and optical indication facilitates the positioning and selection of the product, achieving the purpose of online shopping; the implementation of the present invention will increase merchant income, reduce consumer expenditure, provide jobs, reduce infrastructure pressure, drive the sales of agricultural and sideline products, and boost the offline real economy;

[0090] 2. The transaction collection device of the present invention can complete the collection more quickly than the existing collection equipment such as mobile phones and cameras, saving time; pre-processing protects privacy and increases information security; optical indication facilitates commodity entry, selection and search, improving efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0091] Figure 1 It is a structural schematic diagram of a commodity display and transaction method based on AI recognition technology of the present invention;

[0092] Figure 2 It is a flow chart of a commodity display and transaction method based on AI recognition technology of the present invention;

[0093] Figure 3 It is a structural schematic diagram of a commodity display and transaction device based on AI recognition technology of the present invention.

[0094] Markings in the attached figure: 1. Central processing unit; 2. Memory; 3. Optical sensor component; 4. Communication component; 5. Command control component; 6. Built-in program component; 7. Bus; 8. External client. DETAILED DESCRIPTION

[0095] In order to facilitate understanding of the present invention, the present invention will be described more fully below with reference to the relevant drawings. The present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive.

[0096] like Figures 1 to 3 As shown, a commodity display and transaction method and device based on AI recognition technology include:

[0097] S1. Receive commodity information data collected by a collection device or apparatus through a collection module (the collection device or apparatus includes a camera, a mobile phone, a laser modeling scanner, a computer, a headset, a robot, a barcode scanner, a labeling machine, etc.);

[0098] S2. Perform AI recognition processing on the collected commodity information data through the AI ​​recognition processing module;

[0099] S3, protect the privacy of merchants through the privacy protection module;

[0100] S4. Displaying commodities and trading commodities through the transaction support module;

[0101] Through the collection module, AI recognition and processing module, privacy protection module and transaction support module, complex product information is identified and processed, automatically entered and updated, and the physical information of the product is displayed to consumers. Optical indication facilitates product positioning and selection, and realizes the simultaneous online and offline sales of product information, reducing merchants' time without affecting their offline sales, thereby achieving the goal of moving the market online.

[0102] The collected commodity information data:

[0103] Collect product images, models, voice, text, QR codes and other information through optical sensors, voice input devices, cash register networking, documents and other methods;

[0104] Preferably, the device posture position information of the location of the object is:

[0105] The information of the relative spatial position relationship between the acquisition device and the current item, including the rotation angle, pitch angle, focal length, number and other information of the acquisition device; it is used for the collection and update of product information. If there is no item at the product location, it will be automatically removed from the shelf. If the product is blocked, it will be recorded after a delay. It is used for the purchase of specific products or large-sized products (durian, delicacies, mixed-grade products, etc.). When the buyer selects the product, the optical indicator (such as a laser transmitter) indicates the selected product to the merchant based on the device posture position information of the selected product, and draws a graphic on the product picture to represent the selected product, so as to realize the situation of on-site product selection.

[0106] The AI ​​recognition processing method:

[0107] The commodity information data collected by the collection device is analyzed, identified and matched with the data in the database through AI recognition technology, and the commodities are split and the information is merged to generate commodity display information of the single product. The AI ​​recognition processing of the commodity information collection specifically includes the following:

[0108] Character modeling, model training and model learning, extract and identify the features of various types of product information data collected through relevant recognition algorithms, split the identified products, list the product appearance information in units of single products, identify the graphics, text and audio data related to the products, identify the name, price, origin, production and sales time and other product additional information, and merge the product appearance information with the product additional information to generate the product display information of the single product; the product display information is displayed in the form of images, videos, animations, 3D models, text, voice, etc., and the display information includes: product appearance, name, price, origin, production date, number, AI salesperson, store scene, etc.;

[0109] Processing content and methods: including person recognition, object recognition, image recognition and segmentation, speech recognition, etc.; by performing denoising, color conversion, scaling, cropping, rotation, movement, enhancement, sharpening, feature extraction, frame comparison, segmentation, annotation, extraction of posture and position feature points and other operations on the collected image information, and performing comparison, recognition, classification, evaluation and other processing with the established model library, generate model data, text information, QR code information, product display information, output comparison results, etc.;

[0110] Models include: 3D models or image models such as character models, product models, environment models, facility models, scene models, voice models, AI salesperson models, etc.; establish and train models by: extracting feature points, shapes, colors, textures, reflections, bumps, proportions and other parameters from relevant materials to create model data, and then compare and train with similar model data in large quantities to optimize model data to improve the accuracy of model data and store them in the model library; application tools: such as OpenCV library for preprocessing, enhancement, and recognition, TensorFlow to build deep neural network models for recognition, etc.; perform denoising, enhancement, endpoint detection, training, matching recognition, and result output on voice information to complete voice recognition and result output, such as Google, iFlytek and other voice recognition engines.

[0111] Methods for product splitting and information merging:

[0112] The recorded complex commodity information is split into multiple single commodity information, and matched and merged with commodity-related information data to generate detailed single product display information, which is used to display and trade commodities and publish commodity supply and demand information.

[0113] Merchants review and modify the generated product information; merchants can also manually collect and enter relevant information of the products.

[0114] The method of viewing the actual product:

[0115] Consumers can connect to the merchant's data collection device through the merchant's authorization within a limited time, view the actual product and communicate with the merchant or AI salesperson online. Specifically:

[0116] Consumers connect to the merchant's collection device through the merchant's authorization or set rules. The collection device quickly locates the pre-ordered goods or areas based on the stored goods device posture and position information according to the consumer's needs, and sends back the current image or video of the goods or allows you to view the goods online for a limited time. You can also connect to the merchant or AI salesperson at the same time, and communicate with the consumer through audio and video through voice devices or other terminals. The AI ​​salesperson can automatically answer and sell on behalf of the merchant based on the questions raised by the consumer. The merchant's equipment will emit corresponding sound and light, voice prompts, such as browsing, ordering, etc. according to different stages of the transaction;

[0117] According to the situation of consumers selecting goods on site, select goods such as large or special goods (durian, delicacies, mixed-grade goods, etc.), optical indicators (such as laser transmitters) emit visible light to indicate the goods according to the posture and position information of the goods selected by the buyer, and use the color or stroboscopic light to indicate different states (such as browsing, connecting, placing orders, turning on and off, etc.) to the merchant, and draw graphics on the goods image to indicate the selected goods, so that the merchant can grasp the transaction process and handle it in time; the equipment should be installed on the top (vertical to the goods), outside the stall facing the stall (facing the merchants, with their backs to consumers), etc., to facilitate identification and protection;

[0118] The method of displaying and trading goods:

[0119] Display products: push nearby products or markets based on the product name selected by the consumer and the relative distance between the two parties. The display mode is in the product mode or market mode.

[0120] The specific methods of displaying and trading goods are as follows: agricultural and sideline products have four links: production, wholesale, retail and consumption, involving production groups (such as farmers, agricultural cooperatives, individual industrial and commercial households, fresh food stores, non-staple food stores), wholesale groups (such as vegetable wholesalers, intermediaries, wholesale market practitioners), retail groups (farmers' market stall owners, supermarkets, vegetable stores, fresh food stores, market vendors), consumer groups (such as residents, restaurants, canteens, vegetable processing plants), etc.

[0121] According to the supply and demand relationship of the entire industry chain, the commodity trading model is set as: wholesale and retail model; merchants set the wholesale and retail quantities and corresponding prices according to their own business nature; consumers choose the corresponding purchasing model according to their own characteristics and demand;

[0122] According to the name of the product selected by the consumer and the relative distance between the two parties, the nearby products or markets are pushed. The product display and browsing mode is: Product mode (product plus merchant), which is suitable for consumers to purchase products with clear types and quantities. In this mode, pre-order products are listed and pushed to consumers for consumers to choose from:

[0123] Market mode (market plus merchants) is suitable for consumers who are not sure about the products they want to purchase or buy products randomly. In this mode, nearby merchants are displayed to consumers, and consumers enter the merchant page to purchase products randomly.

[0124] Trading goods: Merchants set wholesale and retail quantities and prices based on the nature of their own business, and consumers choose the corresponding wholesale or retail model based on their own characteristics and demand.

[0125] The privacy protection:

[0126] It is used to modify or block the content that is inconvenient to display for people, objects or areas set by the merchant, and to turn on or off the device at a scheduled time. Specifically:

[0127] Set protection areas, privacy protection content, people or objects, etc.; divide the collected data into display areas and non-display areas, do not record or modify information in non-display areas, distinguish, code or replace non-trading people or objects in the display area, and do not store or display privacy information, protection information, and third-party information; control the power on and off of equipment by identifying the door or the upper and lower locks combined with photosensitive sensors, and voice control power on and off. Merchants have the highest authority and set up multi-level network information security precautions. The operation and protection rules of all equipment and devices are authorized and formulated by merchants. When connecting devices to view products according to the set browsing time limit, method and queuing rules, any equipment and device control instructions not authorized by merchants will not be executed; merchants can set relevant protection rules and parameters through privacy protection modules or external programs (such as device clients, third-party software, etc.).

[0128] Transaction support unit: records and manages the commodity transaction process;

[0129] Client: facilitates information browsing, publishing and communication between both parties of the transaction;

[0130] Traceability credit unit: Pesticides, additives, origins of the product production process, transaction process related information, merchant transaction integrity, service satisfaction, etc. are recorded in graphic and text traceability information to prepare for the retrieval and merging of relevant information during AI recognition processing;

[0131] Database: record, store and call relevant data of the system.

[0132] A commodity display and transaction device based on AI recognition technology, characterized in that it includes a central processing unit 1, a memory 2, an optical sensor component 3, a communication component 4, a command control component 5, a built-in program component 6 and a bus 7, wherein the central processing unit 1, the memory 2, the optical sensor component 3, the communication component 4, the command control component 5 and the built-in program component 6 are all connected to the bus 7;

[0133] The optical sensor component 3 includes a passive type or an active type (a passive optical sensor, such as a CCD / CMOS image sensor, etc., an active optical sensor, such as a laser reflection sensor, a three-dimensional imaging sensor, a light generator, etc.; used to complete product imaging, optical indication, etc.);

[0134] The central processor 1 controls and manages the memory 2, the optical sensor component 3, the communication component 4, the instruction control component 5 and the built-in program component 6. The memory 2 stores data, the optical sensor component 3 scans and records the goods and product information, the communication component 4 transmits data, the instruction control component 5 receives the task instructions sent by the consumer (used by the e-commerce platform to control the equipment to complete scanning, shooting, rotation, focusing, positioning and power on and off, and upload the equipment coordinate encoding of the object by sending task instructions), performs mechanical movement, optical indication, sound and electrical action, the built-in program component 6 processes data and instructions, and the external program sends commands to the device to set or modify the relevant parameters of the device.

[0135] The command control component includes: a servo motor, an optical indicator (such as a laser transmitter, an LED indicator light), a sounder, a motion mechanism, etc. The laser indicator uses visible light to indicate the located goods, indicating different transaction statuses. The sounder sets corresponding prompt sounds for different stages of buyer browsing, guiding, positioning, connecting, placing orders, etc.

[0136] The built-in program component is provided with a built-in program;

[0137] Built-in programs include:

[0138] Information processing: The program pre-processes the protection area, person, anti-theft, identification and other information in the collected data according to the rules;

[0139] Equipment posture and position processing: extracting the rotation angle, pitch angle, focal length, number, center point, feature point and other information of the acquisition equipment device to determine the relative spatial position relationship between the acquisition equipment device and the current object;

[0140] Task control instruction set: formulate dedicated control commands for interaction with various software platforms. By receiving task instructions, control the equipment to perform related actions to complete related tasks or feedback related information (the instruction set includes angle coding, position coding, coordinate coding, scanning instructions, recording instructions, rotation instructions, focusing instructions, power on and off instructions, remote connection instructions, direction instructions, data interface docking and transmission instructions, permission instructions, etc.). The instruction set is compatible with Onvif and other protocols.

[0141] Example

[0142] S1. Information collection. After the merchant has placed the goods, the collection device automatically collects the product information according to the set collection path and area combined with the device posture and position information; during manual collection, the device optical indicator emits an indication light to cooperate with the merchant's manual guidance, and collects according to the guidance direction of the end point of the indicator. During this period, voice can be used for assistance, identification of device commands or product information and corresponding processing. For example, if the device voice turns to command, the device will adjust the direction, if it is product price information, then record the product price information, etc.; the merchant sets the collection refresh time interval or collection area, collection path, connection collection device, privacy protection area, privacy protection person or object, etc. in advance;

[0143] The acquisition module receives the appearance, audio, weight, price list, graphic materials and scenes of the goods collected by the acquisition device through image sensors, model sensors, voice sensors, cash register computers, electronic scales, etc., and shoots, records, scans, etc., to generate product information data in the format of images, audio, video, models, etc. During the acquisition and scanning process, an optical indicator can be turned on to indicate the acquisition status, position, etc., and a speaker can be turned on to emit a corresponding prompt sound;

[0144] S2. Through the AI ​​recognition processing module, the collected and generated multiple types of commodity information data are identified, commodities and non-commodities are identified, the identified commodities are split, and the commodity appearance information is listed in units of single items, and the graphic, text, and audio data related to the commodities are identified to identify the name, price, origin, production and sales time and other commodity additional information, and the commodity appearance information and commodity additional information are combined to generate the commodity display information of the single item. The merchant reviews and publishes the appearance, price and additional information of the commodity display information; the protection information of the non-trading area, people, objects, etc. set by the merchant is not processed;

[0145] S3, privacy protection. The data collected by the equipment needs to be segmented and distinguished between the display area and the non-display area. The non-display area shall not be recorded or modified. The non-trading items, people or living things in the display area shall be distinguished and modified. The display content shall not display the merchant setting information and third-party information other than the transaction. The merchant sets and processes the relevant rules and parameters through the privacy protection module or external program (such as: device client, third-party software, etc.), and sets multi-level network information security precautions. The merchant sets the power on and off time to control the power on and off of the equipment. The equipment power on and off, privacy protection, anti-theft alarm and other operations are controlled according to the identification of business status identification objects such as door opening and closing, lock status, photosensitive sensor, voice control, and the power on and off control command is set with priority level;

[0146] S4. Consumers browse products and nearby products or markets are pushed based on the product names selected by the consumers and the relative distances between the two parties. Logistics and delivery are buyer-centered and mainly based on a short-distance delivery network (same street, district, city, province, etc.). This can be accomplished by third-party companies, merchants delivering goods themselves, neighbors and friends delivering goods for a fee, and approved freelancers or retirees.

[0147] The types and quantities of goods that consumers purchase are relatively clear, and the commodity model (commodity plus merchant) is used to list pre-ordered goods for consumers to choose from;

[0148] When consumers are not sure about the products they are purchasing or they purchase products randomly, the merchant model (merchants plus products) is used to display nearby merchants to consumers. Consumers enter the merchant page and purchase products randomly. Merchants in the same farmers' market are grouped into an online virtual market.

[0149] Consumers connect to the merchant's collection device through the merchant's authorization or set rules, and send back the product status image, video or limited time viewing of the product. They can also view and communicate about the product online. The optical indicator quickly locates the pre-ordered product based on the posture and position data of the product selected by the consumer to indicate the product selected for the order, and uses different flashes or colors to indicate browsing, scanning, ordering and other status;

[0150] Supply and demand information release, producers release information such as the name, specification, sales time, price, output, origin, etc., wholesalers release demand information such as name, quantity, price, etc.; production groups, wholesale groups, retail groups (, consumer groups use the supply and demand information release and communication functions according to their needs, and the system supervises and serves transaction behaviors;

[0151] The consumer selects the goods to be purchased, determines the quantity to be purchased, selects wholesale or retail mode to generate an order and sends it to the merchant for verification. After verification, the merchant measures and prices the goods and sends the graphic and text information of the goods measurement to the consumer for verification. The consumer confirms the measured goods, pays, and the merchant ships the goods. The transaction is completed after receiving the goods.

[0152] It can be assisted by connecting collection devices, electronic scales, cash register computers, label machines, and barcode scanners;

[0153] The commodity trading model is: wholesale and retail model; merchants set the wholesale and retail quantities and corresponding prices according to their own business nature; consumers choose the corresponding purchasing model according to their own characteristics and demand;

[0154] The commodity display and transaction method and device based on AI recognition technology of the present invention are all common mechanical methods for installation, connection or setting. As long as they can achieve their beneficial effects, they can be implemented. Technical personnel in the industry only need to install and operate them according to the accompanying instruction manual, without the need for technical personnel in this field to make creative labor.

[0155] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A commodity display and transaction method based on AI recognition technology, characterized in that: include: S1. Receive commodity information data collected by a collection device or apparatus through a collection module; S2. Perform AI recognition processing on the collected commodity information data through the AI ​​recognition processing module; S3, protect the privacy of merchants through the privacy protection module; S4. Displaying commodities and trading commodities through the transaction support module; The product information data: Receive product appearance, name, price and device posture position information recorded manually and automatically according to time intervals, scanning areas, scanning paths, for input and update of product information and indication and viewing of physical products; Manual recording: The merchant generates a guided scanning path and scanning area by combining the light spot of the optical indicator with the end point of the column to follow the voice and image recognition. When entering the product information, the merchant enters the corresponding voice, picture and text information according to the indication of the product by the optical indicator to complete the collection; The scanning path generates corresponding quantities and numbers according to the merchant's product display and site conditions. The scanning area is divided into display area and non-display area. The display area is mainly used for product information collection and display, and the non-display area is mainly used for privacy protection. The automatic recording: according to the time setting, business status and scanning path, the device is started to automatically complete the collection and update of commodity information according to the scanning path and scanning area; The business status: through business status identification and detection, control the power on and off of equipment and devices to protect privacy information; The AI ​​recognition processing method: The commodity information data collected by the collection device is analyzed, identified and matched with the data in the database through AI recognition technology, and the commodities are split and the information is merged to generate commodity display information of the single product; Methods for product splitting and information merging: The recorded complex commodity information is split into multiple single commodity information, and matched and merged with commodity-related data information to generate detailed single product display information, which is used to display and trade commodities and publish commodity supply and demand information; Instructions and methods of viewing the actual goods: The optical indicator light is used to illuminate the indicated object, and the changing light indicates different information. The constant light indicates that the buyer is selecting the product, and the strobing light indicates that the product has been selected. The consumer draws a graphic on the product image to indicate the selected product. The consumer connects to the merchant's collection device through the merchant to view the actual product and communicate with the merchant online. The privacy protection method described: By identifying objects in business status, equipment is controlled in a timed manner, and privacy information is protected and anti-theft alarms are processed based on the protection rules for areas, people, and objects set by merchants.

2. A commodity display and transaction method based on AI recognition technology as claimed in claim 1, characterized in that: The transaction support module includes: Location distribution unit: pushes goods and matches the best and shortest delivery service in the same area according to the relative locations of the two parties to the transaction, and manages the delivery and receipt of goods; Transaction support unit: records and manages the commodity transaction process; Client: facilitates information browsing, publishing and communication between both parties of the transaction; Traceability credit unit: Pesticides, additives, origins of the product production process, transaction process related information, merchant transaction integrity, and service satisfaction are recorded in graphic and text traceability information to prepare for the retrieval and merging of relevant information during AI recognition processing; Database: record, store and call relevant data of the system.

3. A commodity display and transaction device based on AI recognition technology, which is used to implement the commodity display and transaction method based on AI recognition technology as described in any one of claims 1-2, characterized in that: The device comprises a central processing unit (1), a memory (2), an optical sensor component (3), a communication component (4), a command control component (5), a built-in program component (6) and a bus (7), wherein the central processing unit (1), the memory (2), the optical sensor component (3), the communication component (4), the command control component (5) and the built-in program component (6) are all connected to the bus (7); The central processor (1) controls and manages the memory (2), the optical sensor component (3), the communication component (4), the instruction control component (5) and the built-in program component (6); the memory (2) stores data; the optical sensor component (3) scans and records commodities and commodity information; the communication component (4) transmits data; the instruction control component (5) receives task instructions sent by consumers, performs mechanical movement, optical indication, sound generation and electrical action; and the built-in program component (6) processes data and instructions.

4. A commodity display and transaction device based on AI recognition technology as claimed in claim 3, characterized in that: The optical indication is to emit visible light to the designated object through a light generator, use changing light to represent different information, and complete the scanning with manual guidance.

5. A commodity display and transaction device based on AI recognition technology as claimed in claim 3, characterized in that: The built-in program component is provided with a built-in program; Built-in programs include: Information processing: The program pre-processes the protection area, person, anti-theft, and identification information in the collected data according to the rules, and the control components cooperate to complete manual guidance and following, perform automatic collection and recording, perform identification detection, process data and transmit; Equipment posture and position processing: extract the rotation angle, pitch angle, focal length, number, center point, and feature point information of the relative spatial position relationship between the acquisition equipment and the current object; Task control instruction set: formulate dedicated control commands for interaction between various software platforms. By receiving task instructions, the control device performs related actions to complete related tasks or feedback related information.

Images