Processing method without attaching parts

Abstract

The invention relates to a no-attachment processing method, and relates to the technical field of logistics. The method comprises the following steps of: obtaining a non-attached picture-text pair without an attached piece; the image-text pair without the parcel comprises an image of the parcel without the parcel and the information of the parcel; on the basis of matching processing on the image-text pairs without the express items and a plurality of sample express item image-text pairs in a matching library, determining target matching image-text pairs without the express items from the plurality of sample express item image-text pairs; the matching library comprises sample express item image-text pairs of a plurality of sample express items, and the sample express item image-text pair of each sample express item comprises an express item image and mailing item information of the sample express item; according to the sending information of the target sample express item to which the target matching image-text belongs, determining a target sending record corresponding to the absence of the express item, and according to the target sending record, determining the receiving information of the absence of the express item; and according to the receiving information of the absence of the parcel, carrying out logistics transportation on the absence of the parcel. By adopting the method, the accuracy of logistics transportation without attached pieces can be improved.

Description

Technical Field

[0001] This application relates to the field of logistics technology, and in particular to a method, apparatus, computer equipment, computer-readable storage medium, and computer program product for handling unattached items. Background Technology

[0002] In the logistics industry, undeliverable packages refer to parcels that cannot be delivered to the recipient or returned to the sender. These packages typically have waybills affixed to their outer packaging containing sender and recipient information; however, these waybills often fall off or are damaged during transport, resulting in undeliverable packages. Because undeliverable packages lack necessary logistics information, logistics personnel face difficulties in carrying out subsequent transport operations.

[0003] In related technologies, the logistics information of undeliverable packages is usually traced based on the routing information of the undeliverable packages, such as the sending logistics network, sending information, and receiving information, so as to facilitate the continued logistics transportation operation of the undeliverable packages. However, when faced with complex situations such as missing routing information and a large number of packages, it is difficult to accurately trace the logistics information of undeliverable packages based on the routing information, which leads to a low accuracy rate in the logistics transportation of undeliverable packages. Summary of the Invention

[0004] Therefore, it is necessary to address the aforementioned technical problem of low accuracy in transporting undelivered items by providing a method, apparatus, computer equipment, computer-readable storage medium, and computer program product for handling undelivered items that can improve the accuracy of logistics transportation of undelivered items.

[0005] Firstly, this application provides a method for processing undecorated items, including:

[0006] Obtain the image and text pair of the undeliverable item; the image and text pair of the undeliverable item includes the express image and the consignment information of the undeliverable item;

[0007] Based on the matching process of the image-text pair of the undelivered item and multiple sample express delivery image-text pairs in the matching library, the target matching image-text pair of the undelivered item is determined from the multiple sample express delivery image-text pairs; the matching library includes sample express delivery image-text pairs of multiple sample express delivery items, and each sample express delivery image-text pair includes the express delivery image and the consignment information of the sample express delivery item.

[0008] Based on the shipping information of the target sample parcels to which the target matching image and text pair belongs, determine the target shipping record corresponding to the undeliverable parcel, and based on the target shipping record, determine the recipient information of the undeliverable parcel;

[0009] Based on the receipt information of the unclaimed item, the unclaimed item is transported by logistics.

[0010] Secondly, this application also provides a device for processing unattached parts, comprising:

[0011] The acquisition module is used to acquire the image-text pair of the undelivered item; the image-text pair of the undelivered item includes the express image of the undelivered item and the consignment information;

[0012] The matching module is used to determine the target matching image and text pair for the undelivered item from the multiple sample express item image and text pairs based on the matching processing of the image and text pair of the undelivered item and multiple sample express item image and text pairs in the matching library; the matching library includes sample express item image and text pairs of multiple sample express items, and each sample express item image and text pair of sample express items includes the express item image and the consignment information of the sample express item;

[0013] The determination module is used to determine the target shipment record corresponding to the undeliverable item based on the shipment information of the target sample express item to which the target matching image and text belongs, and to determine the recipient information of the undeliverable item based on the target shipment record;

[0014] The transportation module is used to transport the unattached item according to the receipt information of the unattached item.

[0015] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0016] Obtain the image and text pair of the undeliverable item; the image and text pair of the undeliverable item includes the express image and the consignment information of the undeliverable item;

[0017] Based on the matching process of the image-text pair of the undelivered item and multiple sample express delivery image-text pairs in the matching library, the target matching image-text pair of the undelivered item is determined from the multiple sample express delivery image-text pairs; the matching library includes sample express delivery image-text pairs of multiple sample express delivery items, and each sample express delivery image-text pair includes the express delivery image and the consignment information of the sample express delivery item.

[0018] Based on the shipping information of the target sample parcels to which the target matching image and text pair belongs, determine the target shipping record corresponding to the undeliverable parcel, and based on the target shipping record, determine the recipient information of the undeliverable parcel;

[0019] Based on the receipt information of the unclaimed item, the unclaimed item is transported by logistics.

[0020] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0021] Obtain the image and text pair of the undeliverable item; the image and text pair of the undeliverable item includes the express image and the consignment information of the undeliverable item;

[0022] Based on the matching process of the image-text pair of the undelivered item and multiple sample express delivery image-text pairs in the matching library, the target matching image-text pair of the undelivered item is determined from the multiple sample express delivery image-text pairs; the matching library includes sample express delivery image-text pairs of multiple sample express delivery items, and each sample express delivery image-text pair includes the express delivery image and the consignment information of the sample express delivery item.

[0023] Based on the shipping information of the target sample parcels to which the target matching image and text pair belongs, determine the target shipping record corresponding to the undeliverable parcel, and based on the target shipping record, determine the recipient information of the undeliverable parcel;

[0024] Based on the receipt information of the unclaimed item, the unclaimed item is transported by logistics.

[0025] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0026] Obtain the image and text pair of the undeliverable item; the image and text pair of the undeliverable item includes the express image and the consignment information of the undeliverable item;

[0027] Based on the matching process of the image-text pair of the undelivered item and multiple sample express delivery image-text pairs in the matching library, the target matching image-text pair of the undelivered item is determined from the multiple sample express delivery image-text pairs; the matching library includes sample express delivery image-text pairs of multiple sample express delivery items, and each sample express delivery image-text pair includes the express delivery image and the consignment information of the sample express delivery item.

[0028] Based on the shipping information of the target sample parcels to which the target matching image and text pair belongs, determine the target shipping record corresponding to the undeliverable parcel, and based on the target shipping record, determine the recipient information of the undeliverable parcel;

[0029] Based on the receipt information of the unclaimed item, the unclaimed item is transported by logistics.

[0030] The aforementioned methods, apparatus, computer equipment, computer-readable storage media, and computer program products for processing undeliverable packages all include images of the undeliverable package and information about the items being shipped, as do the image-text pairs of the undeliverable package and the sample express package image-text pairs in the matching database. Therefore, by combining image and text to match the image-text pairs of the undeliverable package and the sample express package image-text pairs, the matching accuracy can be improved, resulting in more accurate target matching image-text pairs. Based on the sender information of the sample express packages to which the more accurate target matching image-text pairs belong, the recipient information of the undeliverable package can be accurately traced, thus enabling further processing of the undeliverable package. Accurate logistics transportation; that is, when the waybill affixed to the outer packaging of a package, which records the recipient's and sender's information, falls off or is damaged, resulting in a lost package that cannot be delivered to the recipient or returned to the sender, the lost package handling method based on the above process, by combining images and text, can trace the recipient's information in complex situations such as missing routing information and a large number of packages, thus improving the accuracy of transporting lost packages. Attached Figure Description

[0031] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 This is a flowchart illustrating a method for processing unattached parts in one embodiment;

[0033] Figure 2 This is a flowchart illustrating the steps in one embodiment of determining the target shipment record corresponding to the undeliverable package based on the shipment information of the target sample express package according to the target matching image and text pair, and determining the recipient information of the undeliverable package based on the target shipment record.

[0034] Figure 3 This is a schematic diagram of the matching process for lost items in one embodiment;

[0035] Figure 4 This is a structural block diagram of a device for processing unattached parts in one embodiment;

[0036] Figure 5 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0037] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0038] It should be noted that the terms "first," "second," etc., used in this application can be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "comprising" and "having," and any variations thereof, used in this application, are intended to cover non-exclusive inclusion. The term "multiple" used in this application refers to two or more. The term "and / or" used in this application refers to one of the embodiments, or any combination of multiple embodiments.

[0039] In the field of logistics technology, for express shipments with damaged or lost waybills, the lack of necessary logistics information makes it impossible to continue logistics transportation. If the logistics information of the express shipment is not traced and transportation is not carried out, the express shipment will be lost, which will lead to customer complaints and claims. However, there is a period of time between the logistics company discovering the loss of the express shipment and the customer's complaint and claim. If the logistics information can be traced and transportation is carried out during this period, it will not lead to customer complaints and claims.

[0040] In related technologies, either single matching methods such as route cross-matching, consignment content matching, and image-text matching are used to match missing packages and trace their logistics information, or matching algorithms based on template matching, feature matching, and deep learning matching are used to match missing packages and trace their logistics information. However, single matching methods often still require manual information summarization and verification when faced with problems such as missing routes, insufficient text information, and numerous and complex images, which still requires a lot of manpower and time. Furthermore, the aforementioned matching algorithms are difficult to simultaneously extract text information, consignment category information, consignment brand information, and various deep semantic information for matching. Therefore, based on these technologies, it is difficult to accurately trace the logistics information of missing packages, which can easily lead to customer complaints and subsequent compensation claims.

[0041] In one exemplary embodiment, such as Figure 1As shown, a method for processing unworn devices is provided. This embodiment illustrates the application of this method to a server, but it is understood that the method can also be applied to a terminal, or to a system including both a server and a terminal, and is implemented through the interaction between the server and the terminal. The server can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services. The terminal can be, but is not limited to, various personal computers, laptops, smartphones, tablets, drones, low-altitude aircraft, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, projection devices, etc. Portable wearable devices can include smartwatches, smart bracelets, head-mounted devices, etc. Head-mounted devices can be virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, etc. In this embodiment, the method includes the following steps:

[0042] Step S102: Obtain the image and text pair of the unattached element.

[0043] Undeliverable packages refer to parcels that cannot be delivered to the recipient or returned to the sender. The outer packaging of parcels usually has a waybill with the recipient's and sender's information affixed to it. The waybill affixed to the parcel often falls off or is damaged during logistics transportation, resulting in undeliverable packages.

[0044] The image-text pair for a parcel includes an image of the parcel and an image of the consignment item. The parcel image includes at least one of an image of the outer packaging and an image of the consignment item; the outer packaging image is an image of the parcel's outer packaging, and the consignment item image is an image of the consignment item itself. The consignment item information is text describing the contents of the consignment item. In specific applications, the consignment item information includes at least one of the following: consignment item category, consignment item description text, and consignment item brand identification information. In specific applications, the number of outer packaging images and consignment item images in the image-text pair can be multiple.

[0045] The number of undeliverable items is at least one, and each undeliverable item corresponds to an undeliverable item image-text pair; the undeliverable item image-text pair includes the express image of the undeliverable item and the consignment information.

[0046] Specifically, when a missing item with a detached or damaged waybill is found in the transit or sorting area, a text-image pair of the missing item will be created and reported; the server will obtain the missing item and its text-image pair reported by the transit or sorting area.

[0047] Step S104: Based on the matching process of the image and text pairs of the undelivered item and multiple sample express delivery image and text pairs in the matching library, the target matching image and text pair of the undelivered item is determined from the multiple sample express delivery image and text pairs.

[0048] The matching database includes multiple sample parcel image-text pairs, with each sample parcel corresponding to one sample parcel image-text pair; the sample parcel image-text pair includes the parcel image and the consignment information of the sample parcel.

[0049] Specifically, for each undeliverable item, the server performs matching processing on the undeliverable item's image and text pair and each sample express item image and text pair in the matching library to determine the target matching degree between each sample express item image and text pair and the undeliverable item image and text pair. Based on the target matching degree of each sample express item image and text pair, the server determines the target matching image and text pair with the undeliverable item from multiple sample express item image and text pairs.

[0050] In practical applications, since the matching process is performed between undelivered image-text pairs and sample express image-text pairs, and both undelivered image-text pairs and sample express image-text pairs include express images and consignment information, the matching process between undelivered image-text pairs and sample express image-text pairs includes at least one of image-to-image matching, text-to-text matching, and image-to-text matching.

[0051] Step S106: Based on the sending information of the target sample express shipment belonging to the target matching image and text pair, determine the target sending record corresponding to the undeliverable shipment, and based on the target sending record, determine the receiving information of the undeliverable shipment.

[0052] Among them, the sender information of the express delivery is used to identify the sender and the sender address of the express delivery.

[0053] Among them, the sender of the sample parcel is the sample sender, and the sample sender is configured with at least one associated parcel logistics network point. The associated parcel logistics network point of the sample sender is the parcel logistics network point commonly used by the sample sender.

[0054] Among them, the recipient information of the express delivery is used to identify the recipient and the recipient address of the express delivery.

[0055] The parcel delivery point stores the delivery record of at least one parcel sent from the parcel delivery point. The parcel delivery record includes the delivery information, the recipient information and the parcel image and text pair. The parcel image and text pair includes the parcel image and the consignment information.

[0056] Specifically, for each undeliverable item, firstly, the server determines the sample sender of the target sample package based on the sender information of the target sample package to which the target matching image and text pair belongs. Since the undeliverable item image and text pair of the undeliverable item matches the target matching image and text pair, the server can regard the sample sender of the target sample package to which the target matching image and text pair belongs as the sender of the undeliverable item.

[0057] Then, the server determines the associated shipping logistics network of the sample sender of the target sample package. Since the server regards the sample sender of the target sample package as a sender without a package, the server can regard the associated shipping logistics network of the sample sender of the target sample package as a shipping logistics network of a sender without a package.

[0058] Next, the server retrieves the corresponding shipping records from the associated shipping network points corresponding to the target sample express, and the shipping records whose shipping information matches the shipping information of the target sample express are used as candidate shipping records. Since the shipping information of the candidate shipping records matches the shipping information of the target sample express, the server can regard the candidate shipping records as the shipping records of the sample sender of the target sample express. Furthermore, since the server regards the sample sender of the target sample express as a sender without a shipment, the server can further regard the candidate shipping records as the shipping records of a sender without a shipment.

[0059] Then, the server retrieves the corresponding shipment record from each candidate shipment record that matches the shipment record of the unclaimed shipment. This record is then used as the target shipment record for the unclaimed shipment. Since the server considers the candidate shipment records as the shipment records of the unclaimed shipment sender, and the shipment record of the target shipment record matches the shipment record of the unclaimed shipment, the server can consider the target shipment record as the shipment record of the unclaimed shipment.

[0060] Then, the server can obtain the recipient information of the undeliverable item based on the recipient information in the target sender record.

[0061] In practical applications, the matching database includes matching data. Each piece of matching data includes the sender information corresponding to a sample express shipment, the sample image and text pair, and the associated sender logistics outlet. Among them, the associated sender logistics outlet corresponding to the sample express shipment is the associated sender logistics outlet of the sample express shipment.

[0062] Step S108: Based on the receipt information of the undeliverable item, carry out logistics transportation for the undeliverable item.

[0063] Specifically, for each undeliverable item, the server continues the logistics transportation of the undeliverable item based on the receipt information of that item.

[0064] In practical applications, the server reprints a shipping label for the undeliverable item based on the sender and recipient information in the target shipment record and affixes it to the surface of the undeliverable item to facilitate its subsequent logistics and transportation.

[0065] In the above-mentioned method for handling undeliverable packages, both the undeliverable package image-text pairs and the sample express package image-text pairs in the matching library include express package images and consignment information. Therefore, by combining image and text to match the undeliverable package image-text pairs and the sample express package image-text pairs, the matching accuracy can be improved, resulting in more accurate target matching image-text pairs. Based on the sender information of the sample express packages to which the more accurate target matching image-text pairs belong, the recipient information of the undeliverable package can be accurately traced, thus enabling accurate logistics transportation of the undeliverable package. In other words, when the waybill affixed to the outer packaging of an express package, which records the recipient and sender information, falls off or is damaged, resulting in an undeliverable package that cannot be delivered to the recipient or returned to the sender, the undeliverable package handling method based on the above process, by combining image and text, can trace the recipient information for undeliverable packages that cannot be delivered to the recipient or returned to the sender under complex circumstances such as missing routing information and a large number of express packages, thus improving the accuracy of transportation of undeliverable packages.

[0066] In one exemplary embodiment, the target matching image-text pair is the sample package image-text pair that best matches the unclaimed item; for example, the server selects the sample package image-text pair with the highest target matching degree as the target matching image-text pair.

[0067] In an exemplary embodiment, the target matching image-text pair is a sample package image-text pair that highly matches the unattached item; for example, the server selects sample package image-text pairs with a target matching degree greater than or equal to a first preset matching degree as the target matching image-text pair, or the server sorts each sample package image-text pair in descending order of the corresponding target matching degree and selects sample package image-text pairs with a corresponding order higher than or equal to the first preset order as the target matching image-text pair, or the server simultaneously selects the target matching image-text pair based on both the first preset matching degree and the first preset order.

[0068] Furthermore, in this embodiment, each sample parcel has a corresponding sample sender, and each sample sender corresponds to at least one sample parcel; that is, each sample sender corresponds to at least one sample parcel image-text pair in the matching database. Therefore, when determining the target matching image-text pair, the server can deduplicate each sample parcel image-text pair according to the corresponding sample sender, retaining only the sample parcel image-text pair with the highest matching degree for each sample sender; then, the server selects the target matching image-text pair from the deduplicated sample parcel image-text pairs.

[0069] In one exemplary embodiment, the server encodes the parcel image and the consignment information based on a multimodal embedding model, and performs matching based on the encoded vectors.

[0070] In one exemplary embodiment, such as Figure 2 As shown, step S106 above, which determines the target shipment record corresponding to the undeliverable package based on the shipment information of the target sample express according to the target matching image and text pair, and determines the recipient information of the undeliverable package based on the target shipment record, specifically includes the following steps:

[0071] Step S202: Based on the sending information of the target sample express, determine the associated sending logistics network of the sample sender of the target sample express.

[0072] Step S204: Based on the sender information and the image-text pair of the target sample parcel, determine the target sender record from the sender records stored in the associated sender logistics network.

[0073] Step S206: Determine the recipient information for the undelivered item based on the recipient information in the target mailing record.

[0074] The shipment information of a parcel includes at least the sender identification information of the sender. In specific applications, the sender identification information is used to identify the sender's identity, which may be the sender's name, username in the logistics system, etc.

[0075] The recipient information of a parcel includes at least the recipient's identification information and the recipient's address. In specific applications, the recipient identification information is used to identify the recipient's identity, which may be the recipient's name, username in the logistics system, etc.

[0076] Specifically, the server determines the sample sender of the target sample express based on the sender identification information in the sender information of the target sample express. Since the image and text pair of the unclaimed item matches the image and text pair of the target sample express (target matching image and text pair), the server can regard the sample sender of the target sample express as the sender of the unclaimed item.

[0077] Then, the server determines the associated shipping logistics network of the sample sender of the target sample package. Since the server regards the sample sender of the target sample package as a sender without a shipment, the server can regard the associated shipping logistics network of the sample sender of the target sample package as an associated shipping logistics network of a sender without a shipment.

[0078] Next, the server retrieves from the associated shipping records stored at the corresponding shipping network points of the target sample express that the matching degree between the sender identification information in the corresponding shipping information and the sender identification information in the shipping information of the target sample express is greater than or equal to a second preset matching degree. These records are then used as candidate shipping records. Since the sender identification information of the candidate shipping records matches the sender identification information corresponding to the target sample express, the server can regard the candidate shipping records as shipping records of the sample sender of the target sample express. Furthermore, since the server regards the sample sender of the target sample express as a sender without a shipment, the server can further regard the candidate shipping records as shipping records of a sender without a shipment.

[0079] Next, the server retrieves from each candidate shipment record the shipment record where the matching degree between the corresponding shipment image and text pair and the unclaimed shipment image and text pair is greater than or equal to a third preset matching degree. This is taken as the target shipment record corresponding to the unclaimed shipment. Since the server regards the candidate shipment records as the shipment records of the unclaimed shipment sender, and the shipment image and text pair of the target shipment record matches the unclaimed shipment image and text pair of the unclaimed shipment, the server can regard the target shipment record as the shipment record of the unclaimed shipment.

[0080] Therefore, the server will treat the recipient identification information and recipient address in the target shipment record as the recipient identification information and recipient address corresponding to the undeliverable shipment.

[0081] In this embodiment, step S108, which involves transporting the undelivered item according to its recipient information, specifically includes the following steps: transporting the undelivered item according to the recipient's identification information and the recipient's address.

[0082] Specifically, the server continues the logistics transportation of the undeliverable item based on the recipient's identification information and the recipient's address.

[0083] In this embodiment, based on the shipping information of the target sample parcel, the server can locate the shipping logistics point of the undeliverable parcel and locate the shipping record of the undeliverable sender from the shipping records stored in the shipping logistics point of the undeliverable parcel; based on the undeliverable parcel image and text pair, it can locate the shipping record of the undeliverable parcel from the shipping record of the undeliverable sender, thereby realizing accurate traceability of the receiving information of the undeliverable parcel, thereby improving the accuracy of transportation of undeliverable parcels.

[0084] In an exemplary embodiment, step S204 above, which determines the target shipment record from the shipment records stored at the associated shipment logistics network point based on the shipment information of the target sample shipment and the image-text pair of the unclaimed shipment, specifically includes the following steps: determining at least one candidate shipment record from the shipment records stored at the associated shipment logistics network point based on the shipment information of the target sample shipment and the shipment information of each shipment record; and determining the target shipment record from the candidate shipment records based on the image-text pair of the unclaimed shipment and the image-text pair of each candidate shipment record.

[0085] Specifically, the server retrieves candidate shipping records from the shipping records stored at the associated shipping network points corresponding to the target sample express shipments, where the matching degree between the sender identification information in the corresponding shipping information and the sender identification information in the shipping information of the target sample express shipment is greater than or equal to a second preset matching degree. Then, the server retrieves the target shipping record from each candidate shipping record, where the matching degree between the corresponding express shipment image and text pair and the undelivered item image and text pair is greater than or equal to a third preset matching degree.

[0086] In this embodiment, based on the sending information of the target sample express, the server can locate the sending record of the sender of the undelivered package from the sending records stored at the logistics network point for undelivered packages; based on the image and text pair of the undelivered package, the server can locate the sending record of the undelivered package from the sending record of the sender of the undelivered package, thereby realizing accurate tracing of the receiving information of the undelivered package and improving the accuracy of transportation of undelivered packages.

[0087] In an exemplary embodiment, step S104 above, based on the matching processing of the image-text pairs of undelivered items and multiple sample express delivery image-text pairs in the matching library, determines the target matching image-text pairs of undelivered items from the multiple sample express delivery image-text pairs. Specifically, this includes the following steps: determining the matching degree between each sample express delivery image-text pair and the undelivered item image-text pair based on the matching processing of the image-text pairs of undelivered items and multiple sample express delivery image-text pairs; performing fusion processing on each matching degree of each sample express delivery image-text pair to obtain the target matching degree between each sample express delivery image-text pair and the undelivered item image-text pair; and determining the target matching image-text pair from the multiple sample express delivery image-text pairs based on the target matching degree of each sample express delivery image-text pair.

[0088] The matching process includes at least one of image-to-image matching, text-to-text matching, and image-to-text matching.

[0089] The matching degree includes at least one of the following: image-to-image matching degree, image-to-text matching degree, and text-to-text matching degree.

[0090] Specifically, for each sample parcel image-text pair, the server performs image-to-image matching and / or text-to-text matching and / or image-to-text matching on the sample parcel image-text pair and the unclaimed parcel image-text pair respectively, to obtain the image-to-image matching degree and / or text-to-text matching degree and / or image-to-text matching degree between the sample parcel image-text pair and the unclaimed parcel image-text pair; then, the server merges the matching degrees of the sample parcel image-text pair to obtain the target matching degree between the sample parcel image-text pair and the unclaimed parcel image-text pair.

[0091] Finally, based on the target matching degree of each sample package image-text pair, the server determines the target matching image-text pair for the unclaimed package from multiple sample package image-text pairs.

[0092] In practical applications, image-to-image matching refers to matching between images of express parcels; text-to-text matching refers to matching between information about the items being shipped; and image-to-text matching refers to matching between images of express parcels and information about the items being shipped.

[0093] In practical applications, for each sample express delivery image-text pair, the server calculates the weighted sum of the matching degrees of each sample express delivery image-text pair to obtain the corresponding target matching degree, or the server calculates the average of the matching degrees of each sample express delivery image-text pair to obtain the corresponding target matching degree.

[0094] In this embodiment, the server can determine the matching degree between sample express delivery image-text pairs and undelivered item image-text pairs from multiple dimensions based on image-image matching, text-text matching, and image-text matching between image-text pairs. In this way, it can accurately determine the target matching image-text pair of the undelivered item image-text pair from multiple sample express delivery image-text pairs.

[0095] In an exemplary embodiment, step S104 above, based on the matching processing of the image-text pairs without a lost item and multiple sample express delivery image-text pairs in the matching library, determines the target matching image-text pairs for the lost item from the multiple sample express delivery image-text pairs. Specifically, this includes the following steps: performing image-to-image matching on the express delivery images in the lost item image-text pairs and the express delivery images in the multiple sample express delivery image-text pairs to obtain the image-to-image matching degree between each sample express delivery image-text pair and the lost item image-text pair; filtering out sample express delivery image-text pairs whose image-to-image matching degree satisfies the first preset condition from the multiple sample express delivery image-text pairs to obtain multiple candidate matching image-text pairs; and performing image-to-image matching on the lost item image-text pairs. Text-to-text matching is performed between the consignment information and the consignment information in multiple candidate matching image-text pairs to obtain the text-to-text matching degree between each candidate matching image-text pair and the unclaimed item image-text pair. Image-to-text matching is then performed between the unclaimed item image-text pair and multiple candidate matching image-text pairs to obtain the image-to-text matching degree between each candidate matching image-text pair and the unclaimed item image-text pair. The image-to-image matching degree, text-to-text matching degree, and image-to-text matching degree of each candidate matching image-text pair are fused to obtain the target matching degree between each candidate matching image-text pair and the unclaimed item image-text pair. Based on the target matching degree of each candidate matching image-text pair, the target matching image-text pair is determined from the multiple candidate matching image-text pairs.

[0096] Specifically, firstly, for each sample parcel image-text pair, the server performs image-to-image matching between the parcel image in the sample parcel image-text pair and the parcel image in the unmarked parcel image-text pair to obtain the image-to-image matching degree between the sample parcel image-text pair and the unmarked parcel image-text pair.

[0097] Next, the server selects multiple candidate matching image-text pairs from the multiple sample image-text pairs based on the image-to-image matching degree of each sample image-text pair. For example, the server filters out sample image-text pairs whose image-to-image matching degree is greater than or equal to a fourth preset matching degree; or, the server sorts the sample image-text pairs in descending order of their image-to-image matching degree and selects sample image-text pairs whose order is higher than or equal to a second preset order as candidate matching image-text pairs; or, the server combines the selection of target matching image-text pairs based on the fourth preset matching degree and the second preset order.

[0098] Then, for each candidate matching image-text pair, the server performs text-to-text matching on the consignment information in the candidate matching image-text pair and the consignment information in the unclaimed image-text pair to obtain the text-to-text matching degree between the candidate matching image-text pair and the unclaimed image-text pair. In addition, the server performs cross-image-text matching on the candidate matching image-text pair and the unclaimed image-text pair to obtain the image-text matching degree between the candidate matching image-text pair and the unclaimed image-text pair. Next, the server merges the matching degrees of the candidate matching image-text pair to obtain the target matching degree between the candidate matching image-text pair and the unclaimed image-text pair.

[0099] Finally, based on the target matching degree of each candidate matching image-text pair, the server determines the target matching image-text pair without the object from multiple candidate matching image-text pairs.

[0100] In practical applications, image-to-image matching refers to matching between images of express parcels; text-to-text matching refers to matching between information about the items being shipped; and image-to-text matching refers to matching between images of express parcels and information about the items being shipped.

[0101] In practical applications, for each candidate matching image-text pair, the server calculates the weighted sum of the matching degrees of the candidate matching image-text pair to obtain the corresponding target matching degree, or the server calculates the average of the matching degrees of the candidate matching image-text pair to obtain the corresponding target matching degree.

[0102] In this embodiment, the server can determine the matching degree between sample express delivery image-text pairs and undelivered item image-text pairs from multiple dimensions based on image-image matching, text-text matching, and image-text matching between image-text pairs. In this way, it can accurately determine the target matching image-text pair of the undelivered item image-text pair from multiple sample express delivery image-text pairs.

[0103] In one exemplary embodiment, the consignment information includes the consignment category, which is the category to which the consignment belongs, determined by the sender or logistics staff from multiple preset item categories. In practical applications, the sender or logistics staff needs to select the corresponding consignment category from multiple preset item categories.

[0104] The text-to-text matching in any of the above embodiments is performed in the following manner: text-to-text matching is performed on the consignment category in the unattached text-image pair and the consignment category in multiple text-image pairs to be matched, so as to obtain the text-to-text matching degree between each text-image pair to be matched and the unattached text-image pair.

[0105] Among them, the image-text pairs to be matched are either sample express delivery image-text pairs or candidate matching image-text pairs.

[0106] Specifically, for each image-text pair to be matched, the server performs text-to-text matching on the consignment category in the image-text pair to be matched and the consignment category in the image-text pair without items, and obtains the text-to-text matching degree between the image-text pair to be matched and the image-text pair without items.

[0107] In this embodiment, the server can determine the target matching image-text pair for the unclaimed item by text-text matching based on the type of consignment.

[0108] In an exemplary embodiment, image-text matching is performed as follows: a first image-text matching is performed on the consignment information in the unclaimed image-text pair and the parcel images in multiple image-text pairs to be matched, to obtain a first image-text matching degree between each image-text pair to be matched and the unclaimed image-text pair; a second image-text matching is performed on the parcel images in the unclaimed image-text pair and the consignment information in multiple image-text pairs to be matched, to obtain a second image-text matching degree between each image-text pair to be matched and the unclaimed image-text pair; and the image-text matching degree of each image-text pair to be matched is obtained based on the first and second image-text matching degrees of each image-text pair to be matched.

[0109] Among them, the image-text pairs to be matched are either sample express delivery image-text pairs or candidate matching image-text pairs.

[0110] Specifically, for each image-text pair to be matched, the server performs image-text matching between the parcel image in the image-text pair to be matched and the consignment information in the unclaimed image-text pair to obtain a first image-text matching degree between the image-text pair to be matched and the unclaimed image-text pair. The server also performs image-text matching between the consignment information in the image-text pair to be matched and the parcel image in the unclaimed image-text pair to obtain a second image-text matching degree between the image-text pair to be matched and the unclaimed image-text pair. Then, the server uses both the first and second image-text matching degrees of the image-text pair to be matched as the image-text matching degree of the image-text pair to be matched.

[0111] In this embodiment, the server can determine the target matching image-text pair for unclaimed items based on the matching between the express package image and the consignment category by cross-image-text matching based on the express package image and the consignment information.

[0112] In one exemplary embodiment, the consignment information includes at least one of consignment description text and consignment brand identification information. The consignment description text is text that details the contents of the consignment. In practical applications, logistics personnel can describe the consignment in words, or input an image of the consignment into a text generation model (e.g., a large language model) to generate the consignment description text. The consignment brand identification information identifies the brand to which the consignment belongs, and may specifically be a brand name, brand icon, etc.

[0113] In this embodiment, the above steps involve performing a first image-text matching on the consignment information in the unclaimed image-text pair and the parcel images in multiple image-text pairs to be matched, and determining the first image-text matching degree between each image-text pair to be matched and the unclaimed image-text pair. Specifically, this includes the following steps: performing a first image-text matching on the consignment description text in the unclaimed image-text pair and the parcel images in multiple sample parcel image-text pairs to obtain the first image-text matching degree between each sample parcel image-text pair and the unclaimed image-text pair.

[0114] Specifically, for each image-text pair to be matched, the server performs image-text matching between the parcel image in the image-text pair to be matched and the consignment description text in the unclaimed image-text pair, and obtains the first image-text matching degree between the image-text pair to be matched and the unclaimed image-text pair.

[0115] In this embodiment, the above steps involve performing a second image-text matching between the parcel image in the undelivered parcel image-text pair and the consignment information in multiple sample parcel image-text pairs to determine the second image-text matching degree between each consignment image-text pair and the undelivered parcel image-text pair. Specifically, this includes the following steps: performing a second image-text matching between the parcel image in the undelivered parcel image-text pair and the consignment brand identification information in multiple sample parcel image-text pairs to obtain the second image-text matching degree between each sample parcel image-text pair and the undelivered parcel image-text pair.

[0116] Specifically, for each image-text pair to be matched, the server performs image-text matching between the consignment brand identification information in the image-text pair to be matched and the express image in the image-text pair without a shipment, and obtains the second image-text matching degree between the image-text pair to be matched and the image-text pair without a shipment.

[0117] In this embodiment, the server can determine the target matching image-text pair for the missing item by cross-image-text matching between the package image and the description text of the item, and between the package image and the brand identification information of the item.

[0118] In an exemplary embodiment, the matching library in step S104 above is constructed in the following manner: based on historical undelivered packages, sample packages are obtained; based on the senders of historical undelivered packages, sample senders of sample packages are obtained; at least one associated shipping logistics network point of the sample sender is determined; sender identification information of the sample sender is obtained to obtain the shipping information of the sample package; based on the package image and consignment information of the sample package, sample package image-text pairs of the sample package are constructed; and a matching library is constructed based on the sample package image-text pairs, shipping information and associated shipping logistics network points corresponding to different sample packages.

[0119] Among them, historical undeliverable items are those whose logistics information has been traced back to undeliverable items.

[0120] Among them, the associated shipping logistics outlets of the sample sender are the shipping logistics outlets that the sample sender frequently uses.

[0121] Specifically, the server collects multiple historical undeliverable packages with traceable logistics information. For each historical undeliverable package, firstly, the server identifies the package as a sample package, the sender as the sample sender, and identifies at least one associated logistics point for the sample sender. Then, the server obtains the sender's identifier information and identifies it as the package's shipping information. Next, the server obtains the package image and consignment information, constructs a sample package image-text pair based on the image and information, and establishes a correspondence between the package's shipping information and the image-text pair. Then, the server combines the package's shipping information, image-text pair, and associated logistics point into matching data, where the associated logistics point is the associated logistics point of the sample sender. Finally, the server combines the matching data for each sample package to obtain a matching database.

[0122] In practical applications, the server manages the matching data in the matching database by grouping them according to the corresponding shipping information. For example, multiple matching data corresponding to the same shipping information can be grouped together for management.

[0123] In this embodiment, the server can build a matching database based on historical missing packages. The matching database includes multiple matching data entries. Each matching data entry includes the sender information, sample image and text pair, and associated sender logistics network points corresponding to a sample package. Based on the matching database, the server can quickly and accurately trace the logistics information of missing packages when they occur. As the number of resolved missing packages increases, the matching database will also be enriched, thereby further optimizing the accuracy of tracing the logistics information of missing packages.

[0124] In an exemplary embodiment, before determining the target matching image-text pair for the undelivered item from the multiple sample express delivery image-text pairs in step S104 above, based on the matching processing of the undelivered item image-text pair and multiple sample express delivery image-text pairs in the matching library, the following steps are further included: performing matching processing on the express delivery image in the undelivered item image-text pair and multiple interference images in the rejection database to determine the similarity between each interference image and the express delivery image in the undelivered item image-text pair; and performing matching processing on the undelivered item image-text pair and multiple sample express delivery image-text pairs in the matching library when the similarity of each interference image is less than the similarity threshold.

[0125] The rejection database includes several interfering images that can create noise interference in tracing the receipt information of undelivered items, such as images that do not contain any information related to the consigned item.

[0126] Specifically, after obtaining the image-text pair of the undelivered item, the server first needs to match the package image in the undelivered item image-text pair with multiple interference images in the rejection database to determine the similarity between each interference image and the package image in the undelivered item image-text pair. Only when the similarity of each interference image is less than the similarity threshold, that is, when the package image in the undelivered item image-text pair does not match any interference image, does the server then match the undelivered item image-text pair with multiple sample package image-text pairs in the matching database.

[0127] If the image of the shipment in the image-text pair of the missing shipment matches any interfering image, the server will not use the missing shipment processing method provided in this application to trace the logistics information of the missing shipment.

[0128] In this embodiment, the server can avoid unnecessary data processing and save server computing resources by pre-determining whether the package image in the undelivered image-text pair is an interfering image.

[0129] To more clearly illustrate the method for processing lost items provided in the embodiments of this application, a specific embodiment is described below. However, it should be understood that the embodiments of this application are not limited thereto. In one exemplary embodiment, this application also provides a method for matching lost items, specifically including the following steps:

[0130] I. Construct a rejection database and a matching database.

[0131] The server collects multiple interfering images without any information related to the shipped items to build a rejection database. The server also collects multiple resolved historical undeliverable packages and identifies the senders, associated logistics outlets of the senders, and the images, categories, descriptions, and brand information of the historical undeliverable packages. Based on these information, the server constructs matching image-text pairs for the historical undeliverable packages. Finally, the server builds a matching database based on the senders, associated logistics outlets, and matching image-text pairs.

[0132] II. Matching process for lost items.

[0133] In this embodiment, lost items refer to parcels with damaged or missing waybills found at transit or sorting areas. For example... Figure 3 The diagram shown is a schematic of the matching process for lost items in this embodiment.

[0134] Step S302: Construct a lost item image-text pair based on the lost item's image, type of item, description text, and brand identification information.

[0135] Step S304: Search the image of the lost package in the rejection database. If an interfering image with a similarity greater than or equal to the preset similarity can be found in the rejection database, it means that the image of the lost package is invalid and no further matching will be performed.

[0136] Specifically, if the lost package image and text pair includes multiple package images, then a search needs to be performed in the rejection database for each package image.

[0137] Step S306: If no interfering image with a similarity greater than or equal to the first preset similarity cannot be found in the rejection database, then the matching database is searched based on the image of the lost package image-text pair to find candidate matching image-text pairs whose first similarity with the lost package image is greater than or equal to the second preset similarity.

[0138] Specifically, if the lost item image-text pair includes multiple package images, then for each package image, it is necessary to determine the search in the matching library. For example, for each package image of the lost item image-text pair, 20 candidate matching image-text pairs with a first similarity of not less than 0.7 are searched.

[0139] Step S308: For each candidate matching image-text pair, calculate the second similarity between the consignment category in the candidate matching image-text pair and the consignment category in the lost item image-text pair.

[0140] Specifically, if the category of the item to be shipped is missing in either the lost item image / text pair or the candidate matching image / text pair, the second similarity is set to 0.5.

[0141] Step S310: For each candidate matching image-text pair, calculate the third similarity between the express delivery image in the candidate matching image-text pair and the item description text in the lost and found image-text pair.

[0142] Specifically, if the description text of the consigned item is missing in the image-text pair of lost items, the third similarity is set to 0.15.

[0143] Step S312: For each candidate matching image-text pair, calculate the fourth similarity between the consignment brand identification information in the candidate matching image-text pair and the express image in the lost and found image-text pair.

[0144] Specifically, if the brand identification information of the consigned item is missing in the candidate matching image and text pair, the fourth similarity is set to 0.15.

[0145] Step S314: For each candidate matching image-text pair, the first similarity, second similarity, third similarity, and fourth similarity of the candidate matching image-text pair are weighted and summed to obtain the total similarity of the candidate matching image-text pair.

[0146] Step S316: Reorder all candidate matching image-text pairs according to the total similarity, and remove duplicates of the corresponding senders according to the candidate matching image-text pairs. Only the candidate matching image-text pairs with the highest total similarity are retained for the same sender.

[0147] Step S318: Extract the top 10 candidate matching image-text pairs from the sorting results as the target matching image-text pairs.

[0148] Step S320: Determine the associated shipping logistics network of the sender corresponding to the target matching image and text pair, search for the shipping record of the sender corresponding to the target matching image and text pair from the shipping records of the associated shipping logistics network, and search for the shipping record that matches the image and text pair of the lost item from the shipping records of the sender corresponding to the target matching image and text pair, thereby tracing the logistics information of the lost item.

[0149] In the field of logistics technology, for express shipments with damaged or lost waybills, the lack of necessary logistics information prevents continued transportation. If the logistics information is not traced and transportation is not carried out, the shipment will be lost, leading to customer complaints and claims. However, there is a period between the logistics company discovering the loss and the customer filing a complaint or claim. If the logistics information can be traced and transportation resumed within this period, customer complaints and claims can be avoided. In this embodiment, a matching database can be used to match lost shipments based on comprehensive images, shipment categories, shipment text descriptions, and shipment brands, thereby improving the accuracy of logistics information tracing and accurately tracing the logistics information of lost shipments for accurate transportation. Automated matching of lost shipments saves manpower and time costs associated with finding lost items, as well as compensation costs. Furthermore, it enables rapid retrieval of lost items, reducing customer complaint risks and improving customer satisfaction.

[0150] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0151] Based on the same inventive concept, this application also provides a device for processing unworn parts to implement the above-described method for processing unworn parts. The solution provided by this device is similar to the solution described in the above method. Therefore, the specific limitations of one or more embodiments of the unworn part processing device provided below can be found in the limitations of the unworn part processing method described above, and will not be repeated here.

[0152] In one exemplary embodiment, such as Figure 4 As shown, a device for processing unattached parts is provided, including: an acquisition module 402, a matching module 404, a determination module 406, and a transport module 408, wherein:

[0153] The acquisition module 402 is used to acquire the image and text pair of the undelivered item; the image and text pair of the undelivered item includes the image of the undelivered item and the consignment information.

[0154] The matching module 404 is used to determine the target matching image and text pair for the undelivered item from the multiple sample express delivery image and text pairs based on the matching processing of the image and text pairs of the undelivered item and the multiple sample express delivery image and text pairs in the matching library; the matching library includes sample express delivery image and text pairs of multiple sample express delivery items, and each sample express delivery image and text pair of sample express delivery items includes the express delivery image of the sample express delivery item and the consignment information.

[0155] The determination module 406 is used to determine the target sending record corresponding to the undeliverable item based on the sending information of the target sample express item to which the target matching image and text belongs, and to determine the receiving information of the undeliverable item based on the target sending record.

[0156] The transportation module 408 is used to transport the undelivered items according to the receipt information of the undelivered items.

[0157] In an exemplary embodiment, the sample sender of the sample parcel is configured with at least one associated parcel delivery logistics point. The associated parcel delivery logistics point stores the delivery records of at least one parcel sent from the associated parcel delivery logistics point. The parcel delivery records include parcel receipt information, delivery information, and parcel image-text pairs. The parcel image-text pairs include at least a parcel image and consignment information.

[0158] The determination module 406 is also used to determine the associated shipping logistics network of the sample sender of the target sample express based on the sender information of the target sample express; the sender information includes at least sender identification information; based on the sender information of the target sample express and the image-text pair of the unclaimed item, the target sender record is determined from each sender record stored in the associated shipping logistics network; the recipient information of the unclaimed item is determined based on the recipient information in the target sender record; the recipient information includes at least recipient identification information and recipient address.

[0159] In an exemplary embodiment, the determining module 406 is further configured to determine at least one candidate shipment record from the shipment records stored at the associated shipment logistics network point based on the shipment information of the target sample shipment and the shipment information of each shipment record; and to determine the target shipment record from the candidate shipment records based on the shipment image-text pair without a shipment and the shipment image-text pair of each candidate shipment record.

[0160] In an exemplary embodiment, the matching module 404 is further configured to determine the matching degree between each sample express delivery image-text pair and the undelivered image-text pair based on the matching processing of the undelivered image-text pair and multiple sample express delivery image-text pairs; the matching processing includes at least one of image-image matching, text-text matching, and image-text matching; the matching degree includes at least one of image-image matching degree corresponding to image-image matching, image-text matching degree corresponding to image-text matching, and text-text matching degree corresponding to text-text matching; the matching degrees of each sample express delivery image-text pair are fused to obtain the target matching degree between each sample express delivery image-text pair and the undelivered image-text pair; based on the target matching degree of each sample express delivery image-text pair, the target matching image-text pair is determined from the multiple sample express delivery image-text pairs.

[0161] In an exemplary embodiment, the matching module 404 is further configured to perform image-to-image matching on the parcel image in the unclaimed parcel image-text pair and the parcel image in multiple sample parcel image-text pairs to obtain the image-to-image matching degree between each sample parcel image-text pair and the unclaimed parcel image-text pair; select sample parcel image-text pairs from the multiple sample parcel image-text pairs whose image-to-image matching degree satisfies the first preset condition to obtain multiple candidate matching image-text pairs; and perform text-to-text matching on the consignment information in the unclaimed parcel image-text pair and the consignment information in the multiple candidate matching image-text pairs to obtain... The text-to-text matching degree between each candidate matching text-image pair and the blank text-image pair is obtained by performing text-to-text matching between the blank text-image pair and multiple candidate matching text-image pairs. The text-to-text matching degree between each candidate matching text-image pair and the blank text-image pair is then fused to obtain the target matching degree between each candidate matching text-image pair and the blank text-image pair. Based on the target matching degree of each candidate matching text-image pair, the target matching text-image pair is determined from the multiple candidate matching text-image pairs.

[0162] In one exemplary embodiment, the consignment information includes the consignment category.

[0163] The matching module 404 is also used to perform text-to-text matching on the consignment category in the undelivered item image-text pair and the consignment category in multiple image-text pairs to be matched, so as to obtain the text-to-text matching degree between each image-text pair to be matched and the undelivered item image-text pair; the image-text pairs to be matched are sample express item image-text pairs or candidate matching image-text pairs.

[0164] In an exemplary embodiment, the matching module 404 is further configured to perform a first image-text matching on the consignment information in the unclaimed image-text pair and the parcel images in multiple image-text pairs to be matched, to obtain a first image-text matching degree between each image-text pair to be matched and the unclaimed image-text pair; the image-text pairs to be matched are sample parcel image-text pairs or candidate matching image-text pairs; perform a second image-text matching on the parcel images in the unclaimed image-text pair and the consignment information in multiple image-text pairs to be matched, to obtain a second image-text matching degree between each image-text pair to be matched and the unclaimed image-text pair; and obtain the image-text matching degree of each image-text pair to be matched based on the first and second image-text matching degrees of each image-text pair to be matched.

[0165] In one exemplary embodiment, the consignment information includes at least one of consignment description text and consignment brand identification information.

[0166] The matching module 404 is also used to perform a first image-text matching on the description text of the consigned item in the unclaimed item image-text pair and the image of the express item in multiple sample express item image-text pairs, so as to obtain the first image-text matching degree between each sample express item image-text pair and the unclaimed item image-text pair.

[0167] The matching module 404 is also used to perform a second image-text matching on the parcel image in the undelivered parcel image-text pair and the consignment brand identification information in multiple sample parcel image-text pairs to obtain the second image-text matching degree between each sample parcel image-text pair and the undelivered parcel image-text pair.

[0168] In an exemplary embodiment, the undelivered package processing device further includes a construction module, configured to: obtain sample packages based on historical undelivered packages; obtain sample senders of sample packages based on senders of historical undelivered packages; determine at least one associated shipping logistics point of the sample sender; acquire sender identification information of the sample sender to obtain shipping information of the sample package; construct sample package image-text pairs based on the package image and consignment information of the sample package; and construct a matching library based on the sample package image-text pairs, shipping information, and associated shipping logistics points corresponding to different sample packages.

[0169] In an exemplary embodiment, the matching module 404 is further configured to perform matching processing on the parcel image in the unmarked image-text pair and multiple interfering images in the rejection database, and determine the similarity between each interfering image and the parcel image in the unmarked image-text pair; and perform matching processing on the unmarked image-text pair and multiple sample parcel image-text pairs in the matching database when the similarity of each interfering image is less than the similarity threshold.

[0170] Each module in the aforementioned device without physical components can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0171] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 5 As shown, this computer device includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores data in a matching library. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a non-volatile processing method.

[0172] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0173] In one exemplary embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0174] In one exemplary embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above-described method embodiments.

[0175] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0176] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0177] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0178] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for processing unattached parts, characterized in that, The method includes: Obtain the image and text pair of the undeliverable item; the image and text pair of the undeliverable item includes the express image and the consignment information of the undeliverable item; Based on the matching process of the image-text pair of the undelivered item and multiple sample express delivery image-text pairs in the matching library, the target matching image-text pair of the undelivered item is determined from the multiple sample express delivery image-text pairs; the matching library includes sample express delivery image-text pairs of multiple sample express delivery items, and each sample express delivery image-text pair includes the express delivery image and the consignment information of the sample express delivery item. Based on the shipping information of the target sample parcels to which the target matching image and text pair belongs, determine the target shipping record corresponding to the undeliverable parcel, and based on the target shipping record, determine the recipient information of the undeliverable parcel; Based on the receipt information of the unclaimed item, the unclaimed item is transported by logistics.

2. The method according to claim 1, characterized in that, The sample sender of the sample parcel is configured with at least one associated parcel delivery logistics point. The associated parcel delivery logistics point stores the delivery record of at least one parcel sent from the associated parcel delivery logistics point. The delivery record of the parcel includes the parcel's recipient information, sender information, and parcel image-text pair. The parcel image-text pair includes at least the parcel image and the consignment information. The step of determining the target shipment record corresponding to the unclaimed item based on the shipment information of the target sample express item to which the target matching image and text belongs, and determining the recipient information of the unclaimed item based on the target shipment record, includes: Based on the shipping information of the target sample parcel, the associated shipping logistics network of the sample sender of the target sample parcel is determined; the shipping information includes at least sender identification information; Based on the shipping information of the target sample parcel and the image-text pair of the undelivered parcel, the target shipping record is determined from each of the shipping records stored in the associated shipping logistics network. Based on the recipient information in the target mailing record, determine the recipient information of the undelivered mail; the recipient information includes at least recipient identification information and recipient address.

3. The method according to claim 2, characterized in that, The process of determining the target shipment record from the shipment records stored at the associated shipping logistics network points based on the shipment information of the target sample shipment and the image-text pair of the unclaimed shipment includes: Based on the shipping information of the target sample parcel and the shipping information of each shipping record, at least one candidate shipping record is determined from each shipping record stored in the associated shipping logistics network. Based on the unattached image-text pair and the express image-text pair of each of the candidate shipment records, the target shipment record is determined from each of the candidate shipment records.

4. The method according to claim 1, characterized in that, The process of matching the image-text pair of the unclaimed item with multiple sample express delivery image-text pairs in the matching library, and determining the target matching image-text pair of the unclaimed item from the multiple sample express delivery image-text pairs, includes: Based on the matching process of the unclaimed item image-text pairs and multiple sample express item image-text pairs, the matching degree between each sample express item image-text pair and the unclaimed item image-text pair is determined; the matching process includes at least one of image-image matching, text-text matching, and image-text matching; the matching degree includes at least one of the image-image matching degree corresponding to the image-image matching, the image-text matching degree corresponding to the image-text matching, and the text-text matching degree corresponding to the text-text matching. The matching degrees of each sample express delivery image-text pair are fused to obtain the target matching degree between each sample express delivery image-text pair and the unattached image-text pair; Based on the target matching degree of each sample express delivery image-text pair, the target matching image-text pair is determined from the multiple sample express delivery image-text pairs.

5. The method according to claim 1, characterized in that, The process of matching the image-text pair of the unclaimed item with multiple sample express delivery image-text pairs in the matching library, and determining the target matching image-text pair of the unclaimed item from the multiple sample express delivery image-text pairs, includes: Perform image-to-image matching on the parcel images in the unattached parcel image-text pairs and the parcel images in the multiple sample parcel image-text pairs to obtain the image-to-image matching degree between each sample parcel image-text pair and the unattached parcel image-text pair. From the multiple sample express parcel image-text pairs, the corresponding image-to-image matching degree of the sample express parcel image-text pairs that meet the first preset condition are selected to obtain multiple candidate matching image-text pairs; Text-to-text matching is performed on the consignment information in the unclaimed item image-text pair and the consignment information in multiple candidate matching image-text pairs to obtain the text-to-text matching degree between each candidate matching image-text pair and the unclaimed item image-text pair. Image-to-text matching is performed on the unclaimed item image-text pair and multiple candidate matching image-text pairs to obtain the image-to-text matching degree between each candidate matching image-text pair and the unclaimed item image-text pair. The image-to-image matching degree, text-to-text matching degree, and image-to-text matching degree of each candidate matching image-to-text pair are fused to obtain the target matching degree between each candidate matching image-to-text pair and the text-to-text pair without any object. Based on the target matching degree of each candidate matching image-text pair, the target matching image-text pair is determined from the plurality of candidate matching image-text pairs.

6. The method according to claim 4 or 5, characterized in that, The consignment information includes the category of the consignment; The text-to-text matching is performed in the following manner: Text-to-text matching is performed on the consignment category in the unclaimed item image-text pair and the consignment category in multiple image-text pairs to be matched to obtain the text-to-text matching degree between each image-text pair to be matched and the unclaimed item image-text pair; the image-text pairs to be matched are sample express item image-text pairs or candidate matching image-text pairs.

7. The method according to claim 4 or 5, characterized in that, The image-text matching is performed in the following way: The first image-text matching is performed on the consignment information in the unclaimed image-text pair and the parcel images in multiple image-text pairs to be matched to obtain the first image-text matching degree between each image-text pair to be matched and the unclaimed image-text pair; the image-text pairs to be matched are sample parcel image-text pairs or candidate matching image-text pairs. A second image-text matching is performed on the parcel image in the unattached image-text pair and the consignment information in multiple image-text pairs to be matched, to obtain the second image-text matching degree between each image-text pair to be matched and the unattached image-text pair; The image-text matching degree of each image-text pair to be matched is obtained based on the first image-text matching degree and the second image-text matching degree of each image-text pair to be matched.

8. The method according to claim 7, characterized in that, The consignment information includes at least one of the consignment description text and consignment brand identification information; The step of performing a first image-text match between the consignment information in the unclaimed image-text pair and the parcel images in multiple image-text pairs to be matched, and determining the first image-text matching degree between each image-text pair to be matched and the unclaimed image-text pair, includes: Perform the first image-text matching on the description text of the consigned item in the unclaimed item image-text pair and the express image in the multiple sample express item image-text pairs to obtain the first image-text matching degree between each sample express item image-text pair and the unclaimed item image-text pair. The step of performing a second image-text matching on the parcel image in the unclaimed image-text pair and the consignment information in multiple image-text pairs to be matched, and determining the second image-text matching degree between each image-text pair to be matched and the unclaimed image-text pair, includes: The second image-text matching is performed on the parcel image in the undelivered parcel image-text pair and the consignment brand identification information in multiple sample parcel image-text pairs to obtain the second image-text matching degree between each sample parcel image-text pair and the undelivered parcel image-text pair.

9. The method according to claim 1, characterized in that, The matching library is constructed in the following way: Based on historical undeliverable packages, sample packages are obtained; based on the senders of the historical undeliverable packages, sample senders of the sample packages are obtained; and at least one associated shipping logistics network point of the sample sender is determined. Obtain the sender identification information of the sample sender to obtain the sending information of the sample express package; Based on the parcel image and consignment information of the sample parcel, a sample parcel image-text pair is constructed; The matching database is constructed based on the sample express image and text pairs corresponding to different sample express items, the sender information, and the associated sender logistics outlets.

10. The method according to claim 1, characterized in that, Before determining the target matching image-text pair for the undelivered item from the multiple sample express delivery image-text pairs in the matching library based on the matching process of the undelivered item image-text pair and the multiple sample express delivery image-text pairs, the method further includes: The similarity between each of the interference images and the express delivery images in the text-image pair without a known object is determined by matching the express delivery images in the text-image pair without a known object with multiple interference images in the rejection database. When the similarity of each of the interfering images is less than the similarity threshold, the image-text pairs without any objects are matched with multiple sample express delivery image-text pairs in the matching library.

Images