A container information sharing system based on intelligent information technology

Abstract

The present application relates to the technical field of supply chain management systems, and in particular to a container information sharing system based on intelligent information technology. The system includes: obtaining a first set of images of the container after loading; performing image recognition processing on the first set of images to identify the container number segment information on the container; associating the container number segment information with the driver's work order information; structuring the container number segment information and the associated work order information to form a comprehensive data set; obtaining a second set of images of the container before unloading; and performing embedded integration of the first set of images and the second set of images to form an image collection, etc. It replaces traditional manual operations, significantly improves efficiency and accuracy, reduces manual errors and information inconsistencies, enhances the transparency and traceability of the supply chain, and through the compilation of archives, all data can be properly preserved to facilitate subsequent verification. At the same time, it reduces delays and risks in multi-level information transmission.

Description

Technical Field

[0001] The present application relates to the technical field of supply chain management systems, and in particular to a container information sharing system based on intelligent information technology. Background Art

[0002] With the rapid development of global logistics and trade, container transportation has played an important role in modern supply chain management. However, traditional container management and information sharing methods have many inefficiencies, such as manual verification, information transmission delays, and data island phenomena. These problems have caused the various links in the transportation process to be disconnected from each other, affecting the overall operational efficiency and information accuracy.

[0003] In traditional container management, drivers usually need to manually take photos of containers and pass these photos through multiple levels of delivery (such as fleets, freight forwarders, customers, etc.) step by step, and finally reach the factory or other links that need to be verified. This model not only requires a lot of manpower to organize and verify information, but also has the risk of information errors, delays and even data loss. These problems become particularly serious when facing large-scale, multi-batch orders, posing huge challenges to the timeliness and accuracy of the supply chain. Summary of the invention

[0004] The present application provides a container information sharing system based on intelligent information technology to solve the above problems.

[0005] In a first aspect, the present application provides a container information sharing system based on intelligent information technology, the system comprising:

[0006] Acquire a first set of images of the container after loading; perform image recognition processing on the first set of images to identify the container segment information on the container; associate the container segment information with the driver's work order information; structure the container segment information and the associated work order information to form a coordinated data set; acquire a second set of images of the container before unloading; perform embedded integration on the first set of images and the second set of images respectively to form an image set; organize the image set and the coordinated data set into a compilation archive; send specific information in the coordinated data set to the object terminal according to the upstream and downstream object types.

[0007] Through the above technical solution, the container images before loading and unloading are collected by mobile terminals, and automatic image recognition is performed to identify key information such as container numbers, eliminating the process of manual sorting and verification by managers, and significantly improving work efficiency. The identified container number is automatically associated with the driver's work order information to form a structured data set, thereby reducing the time for manual data sorting. Automatic identification and data association avoid errors in the manual input process, such as data entry errors or misplaced associations, and ensure the accuracy of information. Through the structured sorting of image and work order information, the problem of inconsistent standards caused by different personnel handling data can be reduced, and the consistency and accuracy of data can be maintained. Through the integration of image collection and data, container information can be updated in a timely manner, and specific information can be automatically sent to the object terminal according to the needs of upstream and downstream objects; this instant data sharing helps to improve the transparency and traceability of the supply chain. The resulting compilation archive can effectively save all relevant data, facilitate subsequent verification, enhance the traceability of information, and avoid the inconvenience caused by information loss. In the traditional model, information needs to be gradually transmitted through multiple levels. Through the automated integration process, the intermediate links and human intervention are reduced, and the data is directly transmitted to the required objects, reducing the risk of delays and errors. The automatic transmission and real-time update of information reduces the retention time in each level, ensuring that the information reaches the downstream links in a timely manner, thereby improving the timeliness of supply chain operations.

[0008] Optionally, obtaining the first set of images of the container after loading and the second set of images of the container before unloading includes: obtaining the specific location of the container and the weather and ambient light at the corresponding location; determining a number of shooting angle parameters of the mobile terminal and shooting light parameters corresponding to each shooting angle according to preset information collection requirements, the specific location of the container and the ambient light conditions; determining a shooting angle limit frame according to the shooting angle parameters; sending the shooting angle parameters, the shooting light parameters and the shooting angle limit frame as shooting requirements to the mobile terminal; capturing images of the container through the mobile terminal; collecting a number of real-shot images of the mobile terminal according to the shooting requirements; parsing each of the real-shot images to determine whether the ghosting state in the real-shot image meets the standard; if the ghosting state does not meet the standard, generating a reshooting request and returning it to the mobile terminal; if the ghosting state meets the standard, storing the real-shot images according to the shooting time node and using them as the first set of images or the second set of images.

[0009] Through the above technical solution, by obtaining the specific location, weather, ambient light and other information of the container, combined with the preset shooting requirements, the best shooting angle and light parameters are generated to ensure that the image can be clearly displayed under various conditions. Under guidance, users can achieve accurate image acquisition through mobile terminals, reducing the steps of manual adjustment and trial and error. Through the setting of the shooting angle limit frame and light parameters, users can be intuitively guided to shoot, ensuring that the mobile terminal is always in the best position when shooting. This effect makes the shooting process easier, and even non-professionals can complete high-quality shooting tasks under the guidance of the system. After the shooting is completed, each image is automatically analyzed, the image's smear status and clarity are judged in real time, and the results are quickly fed back to the user. If the image is found to be unqualified, it will prompt to reshoot. This process greatly reduces the subsequent manual review steps and improves the automation level of image quality control. When the image quality is detected to be unqualified (such as smear, blur, etc.), the system automatically issues a reshoot request and guides the operator to readjust the shooting angle, light and other parameters for a second shoot until the image meets the standard. This function improves the reliability of the entire image acquisition process and the consistency of image quality. Once the images meet the standards, the system will automatically classify them as "first set of images" or "second set of images" according to the shooting time node (such as before loading or unloading), and associate them with container numbers, work order information, etc., and store them in the database. This operation ensures that the data is organized and easy to retrieve, and that the images are consistent with the corresponding work orders, container numbers and other information.

[0010] Optionally, collecting a number of real-shot images of the mobile terminal according to the shooting requirements includes: acquiring the real-shot images from the mobile terminal; parsing the real-shot images to obtain the container outline in the image; calculating the overall deviation degree between the container outline and the shooting angle limiting frame; judging whether the overall deviation degree is less than a preset deviation range; if the overall deviation degree is less than the preset deviation range, judging that the container outline meets the shooting angle limiting frame in the shooting requirements; collecting the real-shot images; if the overall deviation degree is greater than the preset deviation range, judging that the container outline does not meet the shooting angle limiting frame in the shooting requirements; and generating a reshooting request and sending it back to the mobile terminal.

[0011] Through the above technical solution, the outline of the container in the image is automatically parsed, and the degree of deviation between it and the shooting angle limit frame is calculated. This automated process reduces manual intervention and speeds up the shooting and review process, ensuring the accuracy and quality of the image, while also improving the speed and efficiency of image analysis. After the shooting is completed, the quality of the image is immediately fed back to determine whether it meets the preset shooting requirements, including angle deviation and image clarity (smear state). This effect ensures that the image quality is controlled during the shooting stage, avoiding rework due to unqualified images during subsequent image processing and analysis. If the deviation of the image exceeds the preset allowable range, the system will immediately generate a reshoot request to notify the operator to readjust the angle and shoot. The reshoot mechanism ensures that each collected image meets the quality standards and the final generated image can meet the requirements of subsequent analysis and archiving.

[0012] Optionally, the calculating the overall deviation degree between the container outline and the shooting angle limiting frame includes: parsing the container outline, extracting the container outline into a plurality of contour lines; parsing the shooting angle limiting frame, extracting the shooting angle limiting frame into limiting frame lines having the same number as the contour lines; analyzing the similarity relationship between the container outline and the shooting angle limiting frame; binding a plurality of contour lines to corresponding limiting frame lines respectively according to the similarity relationship; calculating the corresponding line deviation degree between each contour line and the corresponding limiting frame line; and determining the overall deviation degree between the container outline and the shooting angle limiting frame according to the corresponding line deviation degree.

[0013] Through the above technical solution, by comparing the geometric features of the container's outline and the shooting angle limit frame, it is possible to accurately determine whether the position and angle of the container in the image meet the predetermined shooting requirements. This analysis ensures the accuracy of the image shooting. The deviation between the container outline and the shooting angle limit frame is automatically detected, and efficient and accurate judgment can be completed without human intervention. By calculating the differences in geometric features in multiple dimensions such as position, angle, and length, it is determined whether the image is qualified. After the shooting is completed, the deviation of the container image can be immediately fed back. If the deviation exceeds the preset range, the system will generate a reshoot reminder to notify the user to adjust the shooting angle or position. By analyzing the geometric features of the container outline and the shooting angle limit frame, the shooting quality of each image can be ensured to be consistent. Each image undergoes the same quality inspection and deviation evaluation to ensure the consistency of the image.

[0014] Optionally, the calculation of the deviation degree between each contour line and the corresponding limiting frame line is performed according to the following formula:

[0015] ;

[0016] in, For the The degree of deviation between the contour line and the corresponding line of the corresponding limiting frame line; For the contour lines; For the The limit frame line corresponding to the contour line; For the The directional similarity weight between the contour line and the corresponding limiting frame line; For the The position similarity weight of each contour line and the corresponding limiting frame line; For the The maximum distance between a contour line and the corresponding limit frame line;

[0017] No. The maximum distance between a contour line and the corresponding limit frame line is calculated according to the following formula:

[0018] ;

[0019] in, For the The maximum distance between a contour line and the corresponding limit frame line; is the maximum value function; ; ; For the The endpoints of the contour lines To the corresponding limit box distance; For the The endpoints of the contour lines To the corresponding limit box distance; For the The contour lines correspond to the endpoints of the bounding box lines To Contour lines distance; For the The contour lines correspond to the endpoints of the bounding box lines To Contour lines distance.

[0020] Through the above technical solution, by calculating the maximum distance between the endpoints of the contour line and the limiting frame line, the geometric difference between the two line segments can be accurately quantified. By calculating the directional similarity and position similarity deviations, it is possible to automatically detect whether the contour of the container in each image meets the preset standards of the limiting frame, reducing the necessity of manual intervention. Improve work efficiency and avoid unnecessary subjective errors. By calculating the deviation between all contour lines and the limiting frame line, it is ensured that the quality of each captured image reaches a consistent high standard. This method reduces the quality inconsistency caused by factors such as shooting angle and position.

[0021] Optionally, the sending of specific information in the coordinated data set to the object terminal according to the upstream and downstream object types includes: parsing each of the upstream and downstream object types to determine the corresponding business needs; filtering out demand-related information corresponding to the upstream and downstream objects from the coordinated data set according to the business needs of each upstream and downstream object; matching the filtered demand-related information with the terminals of each object; and determining appropriate communication channels according to different upstream and downstream object types and automatically sending demand-related information.

[0022] Optionally, the first set of images and the second set of images are respectively embedded and integrated to form an image set, including: classifying the images in each set of images to determine a number of information representative images and a number of status representative images; parsing each of the information representative images to determine the text area; cropping each of the information representative images according to the text area and a first preset border range; performing deformity correction on the cropped images to obtain a number of corrected images; integrating a number of similar corrected images in the same set of images to obtain an information card image; parsing the free area in each of the status representative images, embedding the corresponding information card image into the free area to form a combined image paste; and arranging and summarizing the combined image pastes to form an image set.

[0023] Optionally, the parsing of the free area in each of the state representative images and embedding the corresponding information card image into the free area to form a combined image paste includes: parsing each of the state representative images to determine the state representative image resolution, the main outline and extension direction of the container in the image; according to the extension direction, determining that the free corners are different from the extension direction; according to the main outline and the free corners, determining the free area of ​​the largest triangle at the corner; parsing the information card image to obtain the information card image resolution and the minimum character size; according to the state representative image resolution, determining the minimum character display size of the state representative image; adjusting the information card image resolution according to the minimum character display size and the minimum character size; embedding the information card image with adjusted resolution into the free area of ​​the largest triangle at the corner of the state representative image.

[0024] Optionally, the deformity correction of the cropped image to obtain a number of corrected images includes: analyzing the cropped image to determine the light and shadow structure of the cropped image; determining the corrugated distribution structure of the surface of the container according to the light and shadow structure; segmenting the image according to the corrugated distribution structure, dividing the segmented parts into a corrugated flat surface image and a corrugated oblique surface image, and numbering them according to the composition state; determining the segmentation association points according to the position of the segmentation line between the corrugated flat surface image and the corrugated oblique surface image; performing perspective correction on the corrugated flat surface image to obtain a corrugated flat surface front view image; analyzing the projection relationship between the corrugated flat surface image and the corrugated oblique surface image; correcting the corrugated oblique surface image according to the corrugated flat surface front view image and the projection relationship to obtain a corrugated oblique surface front view image; merging the flat surface image with the corrugated oblique surface front view image according to the numbering and the segmentation association points to obtain a corrected image.

[0025] Optionally, the step of integrating several similar corrected images in the same set of images to obtain an information card image includes: identifying the text range of each of the similar corrected images; performing hue weakening and semi-transparency processing on each of the similar corrected images to obtain an overlay base map; overlaying and integrating several overlay base maps according to the text range, performing edge clearing processing on the overlayed and integrated image to obtain an easily recognizable image; and cropping the easily recognizable image according to a second preset border range to obtain an information card image. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0027] Figure 1 A schematic diagram of an application scenario provided for an embodiment of the present application;

[0028] Figure 2 A flowchart of a container information sharing system based on intelligent information technology provided in one embodiment of the present application. DETAILED DESCRIPTION

[0029] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0030] In addition, the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article, unless otherwise specified, generally means that the associated objects before and after are in an "or" relationship.

[0031] The embodiments of the present application are further described in detail below in conjunction with the drawings in the specification.

[0032] In the traditional container management model, drivers usually need to manually take photos of containers and pass these photos through multiple levels of delivery (such as fleets, freight forwarders, customers, etc.) step by step, and finally reach the factory or other downstream links that need to be verified. This model not only requires a lot of manpower to organize and verify information, but also has the risk of information errors, delays and even data loss. These problems become particularly serious when facing large-scale, multi-batch orders, posing huge challenges to the timeliness and accuracy of the supply chain.

[0033] Based on this, the present application provides a container information sharing system based on intelligent information technology. The container images before loading and unloading are collected by mobile terminals, and automatic image recognition is performed to identify key information such as container numbers, eliminating the process of manual sorting and verification by managers, and significantly improving work efficiency. The identified container number is automatically associated with the driver's work order information, and a structured data set is formed, thereby reducing the time for manual data sorting. Automated identification and data association avoid errors in the manual input process, such as data entry errors or misplaced associations, and ensure the accuracy of information. By structuring the image and work order information, the problem of inconsistent standards caused by different personnel processing data can be reduced, and the consistency and accuracy of the data can be maintained. Through the integration of image collection and data, container information can be updated in a timely manner, and specific information can be automatically sent to the object terminal according to the needs of upstream and downstream objects; this instant data sharing helps to improve the transparency and traceability of the supply chain. The formed compilation archive can effectively save all relevant data, facilitate subsequent verification, enhance the traceability of information, and avoid the inconvenience caused by information loss. In the traditional model, information needs to be gradually transmitted through multiple levels. Through the automated integration process, the intermediate links and human intervention are reduced, and the data is directly transmitted to the required objects, reducing the risk of delays and errors. The automatic transmission and real-time update of information reduces the retention time in each level, ensuring that the information reaches the downstream links in a timely manner, thereby improving the timeliness of supply chain operations.

[0034] Figure 1 A schematic diagram of an application scenario provided by the present application. In the container management mode, manual operation and multi-level transmission methods bring many problems. The system provided by the present application is applied. Specifically, the system provided by the present application is applied to any server, and the server interacts with the mobile terminal and the object terminal to optimize container management through automatic image recognition and data processing. First, the container image before loading and unloading is collected by the mobile terminal, and the container number is automatically identified and associated with the driver's work order information to generate a structured data set. This process replaces the traditional manual operation, significantly improves efficiency and accuracy, and reduces the problems of manual errors and inconsistent information. Automatically transmit key data to upstream and downstream to enhance the transparency and traceability of the supply chain. By compiling archives, all data can be properly stored, which is convenient for subsequent verification and avoids information loss. At the same time, delays and risks in multi-level information transmission are reduced, and the timeliness and reliability of the supply chain are improved. The specific implementation method can refer to the following embodiments.

[0035] Figure 2 This is a flow chart of a container information sharing system based on intelligent information technology provided by an embodiment of the present application. The system of this embodiment can be applied to the server in the above scenario. Figure 2 As shown, the system comprises:

[0036] S201, obtaining a first set of images of a loaded container; performing image recognition processing on the first set of images to identify container number segment information on the container.

[0037] The first set of images may be images taken of the container after loading, and the images may include key information such as the container number and appearance.

[0038] The container number segment information may be unique identification information for identifying each container.

[0039] Specifically, in the process of supply chain management, the container management mode often requires manual collection and verification of a large number of images. This operation mode is prone to information errors in heavy and repetitive work, and the manual verification process is very time-consuming, resulting in work delays. Therefore, the container images after loading are collected through mobile terminals as the first set of images; the collected images are denoised and image enhancement technology is used to automatically identify the container segment information in the image using OCR (Optical Character Recognition) technology.

[0040] S202: Associating the container number segment information with the driver's work order information.

[0041] Work order information can be detailed information of the transportation task, such as driver identity, vehicle information, transportation route, order number, loading and unloading time and location, etc.

[0042] Specifically, each work order usually has a unique order number. The order number in the driver's work order is associated with the container number segment as two primary keys. Then, the timestamp identifying the container number segment is compared with the time in the work order through the loading time and unloading time recorded in the work order to ensure that the container number segment is consistent with the time range of the corresponding work order, and the association is verified to be correct once. If multiple orders are carried out at the same time, further matching is required based on the driver's identity (such as the driver's work number or license plate number). The relevant container number segment is associated with the work order information through the driver's identity proof (such as work number, name or license plate number). The vehicle loaded with the container can also be identified through automatic identification technology (such as vehicle license plate recognition), and the vehicle information can be associated with the relevant records in the work order.

[0043] S203, structuring the container number segment information and the associated work order information to form a coordinated data set;

[0044] The data set may be a collection of data generated during the container management process.

[0045] Specifically, the container number segment information is integrated with all relevant data in the work order (such as driver identity, license plate number, loading and unloading location, time, etc.) to generate a complete and structured integrated data set. Through cross-validation of multiple data fields (such as driver identity, vehicle, loading and unloading time and location, etc.), abnormal information will be marked and triggered manual verification.

[0046] S204: Acquire a second set of images of the container before unloading.

[0047] The second set of images can be images of the container taken before unloading. The images can include the state of the container during transportation, whether it is damaged, etc.

[0048] Specifically, when collecting images of the container by a mobile terminal before unloading, the second set of images can be obtained by using the same acquisition system as the first set of images.

[0049] S205, performing embedded integration on the first set of images and the second set of images respectively to form an image set;

[0050] Embedded integration can be an operating system that embeds key information (such as container number segments, status information, etc.) into specific areas of the image during image processing so that these important information can be seen simultaneously when viewing the image. Image collection can be a combination of multiple processed images.

[0051] Specifically, according to preset rules, representative images of key information (such as container number segment, transportation status, etc.) are extracted from the collected images; the physical status of the container (such as container appearance, location, etc.) is classified to generate different status images; each image is distorted or perspective corrected to ensure that the container outline and information area can be correctly displayed; the image area of ​​key information such as the container number segment is accurately captured to extract the container number segment and related work order information; by analyzing the free areas in the container status image (such as the corners or edges of the container), a suitable place for embedding is found; the processed information card image is embedded in the free area of ​​the status image to ensure that the information card does not cover the key visual information of the container.

[0052] S206. Compile the image collection and the overall data set into a compilation archive.

[0053] A compilation archive can be a complete record that combines image collections and coordinated data sets for supply chain tracking and management.

[0054] Specifically, the image collection consists of container images before loading and unloading, which have been embedded and integrated, including clearly visible container number segments, status information, etc. The images are marked according to different categories (such as loading images, unloading images, abnormal images, etc.) so that they can be organized by category when the archive is compiled; the overall data set includes all structured data related to the container; this data has been associated with the container images through previous steps to ensure consistency and integrity. The image collection is integrated with each data item in the overall data set to generate a compilation archive containing image and text data.

[0055] S207. Send specific information in the coordinated data set to the object terminal according to the upstream and downstream object types.

[0056] Specific information can be information related to upstream and downstream objects that is selected from the coordinated data set based on their needs. Object terminals can be terminal devices used by participants in the supply chain.

[0057] Specifically, the type of information required by each object is determined based on its business needs. For example, upstream objects need to pay attention to the loading status and transportation arrangements of orders, while downstream objects are more concerned about the arrival and quality status of goods. According to the needs of each object, filter out relevant information from the coordinated data set. For example, upstream objects may need to obtain the container status and transportation progress before loading, while downstream objects may need container acceptance reports after unloading. Automatically select the appropriate delivery channel based on each object's communication habits and business needs. For example, important cargo status updates can be sent via email or SMS, while daily data can be transmitted to the object's system through an API interface. For ordinary information that does not need to be updated in real time (such as daily reports, regular status updates, etc.), batch push is performed at preset time intervals.

[0058] Through the method provided in this embodiment, the container images before loading and unloading are collected by the mobile terminal, and automatic image recognition is performed to identify key information such as the container number, eliminating the process of manual sorting and verification by the management personnel, and significantly improving work efficiency. The identified container number is automatically associated with the driver's work order information, and a structured data set is formed, thereby reducing the time for manual data sorting. Automatic identification and data association avoid errors in the manual input process, such as data entry errors or misplaced associations, and ensure the accuracy of information. Through the structured sorting of image and work order information, the problem of inconsistent standards caused by different personnel processing data can be reduced, and the consistency and accuracy of the data can be maintained. Through the integration of image collection and data, the container information can be updated in time, and specific information can be automatically sent to the object terminal according to the needs of upstream and downstream objects; this instant data sharing helps to improve the transparency and traceability of the supply chain. The formed compilation archive can effectively save all relevant data, facilitate subsequent verification, enhance the traceability of information, and avoid the inconvenience caused by information loss. In the traditional model, information needs to be gradually transmitted through multiple levels. Through the automated integration process, the intermediate links and human intervention are reduced, and the data is directly transmitted to the required objects, reducing the risk of delays and errors. The automatic transmission and real-time update of information reduces the retention time in each level, ensuring that the information reaches the downstream links in a timely manner, thereby improving the timeliness of supply chain operations.

[0059] In some embodiments, the specific location of the container and the weather and ambient light at the corresponding location are obtained; according to the preset information collection requirements, the specific location of the container and the ambient light conditions, several shooting angle parameters of the mobile terminal and the shooting light parameters corresponding to each shooting angle are determined; according to the shooting angle parameters, the shooting angle limit frame is determined; the shooting angle parameters, the shooting light parameters and the shooting angle limit frame are sent to the mobile terminal as shooting requirements; the container is imaged through the mobile terminal; according to the shooting requirements, several real-shot images of the mobile terminal are collected; each real-shot image is analyzed to determine whether the ghosting state in the real-shot image meets the standard; if the ghosting state does not meet the standard, a re-shooting request is generated and sent back to the mobile terminal; if the ghosting state meets the standard, the real-shot images are stored according to the shooting time node and used as the first set of images or the second set of images.

[0060] Information collection requirements may be standards and conditions set for capturing images, including shooting angles, lighting conditions, image clarity, etc. Shooting angle parameters may be specific angle settings for guiding shooting, used to ensure that key parts of the container, such as container number segments, seal positions, etc., can be covered during shooting. Shooting light parameters may be light-related parameters set during the shooting process, such as exposure time, ISO sensitivity, aperture size, etc. The shooting angle limit frame may be a virtual frame displayed on a mobile terminal to guide the operator to aim at the container for shooting. Actual images may be images actually captured by a mobile terminal, which record the status of the container at different locations and at different points in time. The smear state may be blurring or ghosting in the captured image due to movement or other reasons.

[0061] Specifically, according to the intensity of ambient light (daytime, nighttime or in the shadow), set appropriate exposure time and sensitivity value to ensure the clarity of the image; preset some optimal shooting angles to ensure that the key parts of the container can be covered, especially the container number segment area; according to the specific location of the container and shooting requirements, select the best shooting position of the mobile terminal, display a virtual shooting limit frame on the mobile terminal, guide the user to place the container in the shooting field of view, and ensure that the shooting angle meets the requirements; according to the ambient light conditions, the shooting parameters of the mobile terminal are specifically required, such as exposure time, sensitivity value and aperture size, to ensure that the image is still clearly visible even under poor light conditions; according to the location, shooting angle and light requirements of the container, generate specific shooting instructions, including: recommended shooting angle parameters, light adjustment parameters, and display position of the angle limit frame. After each shooting, the image is transmitted back to the server and an instant preview is displayed on the mobile terminal. Detect whether the image meets the shooting requirements, automatically analyze the shot image through the algorithm, and detect whether there are problems such as blurring and ghosting; if ghosting or other problems affect the image quality, the system will automatically issue a reshoot request and prompt the operator to reshoot through the mobile terminal. After receiving the retake instruction, the operator can adjust the shooting angle or position and take a new round of shots. According to the shooting time node, the images that meet the standards are classified as "first set of images" (after loading) or "second set of images" (before unloading). Each set of images is associated with data such as container number and work order.

[0062] Through the method provided in this embodiment, by obtaining the specific location, weather, ambient light and other information of the container, combined with the preset shooting requirements, the best shooting angle and light parameters are generated to ensure that the image can be clearly displayed under various conditions. Under guidance, the user can achieve accurate image acquisition through the mobile terminal, reducing the steps of manual adjustment and trial and error. Through the setting of the shooting angle limit frame and light parameters, the user can be intuitively guided to shoot, ensuring that the mobile terminal is always in the best position when shooting. This effect makes the shooting process simpler, and even non-professionals can complete high-quality shooting tasks under the guidance of the system. After the shooting is completed, each image is automatically parsed, the image's smear state and clarity are judged in real time, and the results are quickly fed back to the user. If it is found that the image does not meet the requirements, it will prompt to reshoot. This process greatly reduces the subsequent manual review steps and improves the automation level of image quality control. When the image quality is detected to be unqualified (such as smear, blur, etc.), the system automatically issues a reshoot request and guides the operator to readjust the shooting angle, light and other parameters for a second shoot until the image meets the standard. This function improves the reliability of the entire image acquisition process and the consistency of image quality. Once the images meet the standards, the system will automatically classify them as "first set of images" or "second set of images" according to the shooting time node (such as before loading or unloading), and associate them with container numbers, work order information, etc., and store them in the database. This operation ensures that the data is organized and easy to retrieve, and that the images are consistent with the corresponding work orders, container numbers and other information.

[0063] In some embodiments, a real-shot image is acquired from a mobile terminal; the real-shot image is parsed to obtain a container outline in the image; the overall deviation between the container outline and the shooting angle limit frame is calculated; it is determined whether the overall deviation is less than a preset deviation range; if the overall deviation is less than the preset deviation range, it is determined that the container outline meets the shooting angle limit frame in the shooting requirements; the real-shot image is collected; if the overall deviation is greater than the preset deviation range, it is determined that the container outline does not meet the shooting angle limit frame in the shooting requirements; a reshoot request is generated and sent back to the mobile terminal.

[0064] The container outline may be the container edge and shape information extracted from the real-shot image by an image analysis algorithm. The overall deviation degree may be the position and angle deviation between the container outline and the shooting angle limit frame preset by the system. The preset deviation range may be an allowable error range set when measuring the overall deviation degree, and may be stored in a preset database.

[0065] Specifically, the images taken by the operator are received through the mobile terminal, and the image recognition algorithm (such as edge detection, shape analysis, etc.) is used on the received real-shot images to identify and extract the container contour information in the image. The overall deviation degree between the container contour and the shooting angle limit frame is calculated, such as comparing the relative position of the container contour identified in the real-shot image and the shooting angle limit frame, and calculating the overall deviation between the two, including indicators such as angle and position deviation. Based on the calculated overall deviation degree, the system determines whether it is within the preset deviation range. If the actual deviation is less than the range, it is judged that the container position meets the shooting requirements; if it is greater than the range, it is considered that the shooting angle is incorrect and needs to be re-shot. For images that meet the shooting requirements, the system classifies and saves them as the first set of images or the second set of images for subsequent use.

[0066] Through the method provided in this embodiment, the outline of the container in the image is automatically parsed, and the degree of deviation between it and the shooting angle limit frame is calculated. This automated process reduces manual intervention and speeds up the shooting and review process, ensuring the accuracy and quality of the image, while also improving the speed and efficiency of image analysis. After the shooting is completed, the quality of the image is immediately fed back to determine whether it meets the preset shooting requirements, including angle deviation and image clarity (smear state). This effect ensures that the image quality is controlled during the shooting stage, avoiding rework due to unqualified images during subsequent image processing and analysis. If the deviation of the image exceeds the preset allowable range, the system will immediately generate a reshoot request to notify the operator to readjust the angle and shoot. The reshoot mechanism ensures that each collected image meets the quality standards, and the final generated image can meet the requirements of subsequent analysis and archiving.

[0067] In some embodiments, the container contour is parsed and the container contour is extracted into a plurality of contour lines; the shooting angle limiting frame is parsed and the shooting angle limiting frame is extracted into limiting frame lines having the same number as the contour lines; the similarity relationship between the container contour and the shooting angle limiting frame is analyzed; based on the similarity relationship, a plurality of contour lines are respectively bound to corresponding limiting frame lines; the corresponding line deviation degree between each contour line and the corresponding limiting frame line is calculated; and based on the corresponding line deviation degree, the overall deviation degree between the container contour and the shooting angle limiting frame is determined.

[0068] The contour line may be an outer edge line of a container extracted from a real shot image by image processing technology. The limiting frame line may be an edge line of a shooting reference frame extracted from a shooting reference frame by image processing technology. The similarity relationship may be a geometric similarity between the container contour line and the shooting angle limiting frame line. The line deviation degree may be a geometric difference between the container contour line and the corresponding shooting angle limiting frame line.

[0069] Specifically, an edge detection algorithm is used to extract the outer boundary of the container from the image as the extracted contour line of the container, and a set of continuous edge points are fitted into a straight line using methods such as Hough transform or least squares method, thus abstracted into several line segments. The shooting angle limit frame is parsed into several limit frame lines by the above means, and the number is the same as the container contour line. Through the minimum distance matching algorithm, the similarity relationship is determined, and each contour line of the container is bound to the corresponding shooting angle limit frame line. For each pair of bound contour lines and limit frame lines, the degree of corresponding line deviation between them is calculated. The deviation can be calculated in multiple dimensions of position and angle; based on the corresponding line deviation between all contour lines and limit frame lines, the overall deviation degree is calculated as the basis for evaluating whether the image meets the shooting requirements.

[0070] Through the method provided by this embodiment, by comparing the geometric features of the container's outline and the shooting angle limiting frame, it is possible to accurately determine whether the position and angle of the container in the image meet the predetermined shooting requirements. This analysis ensures the accuracy of the image shooting. The deviation between the container outline and the shooting angle limiting frame is automatically detected, and efficient and accurate judgment can be completed without human intervention. By calculating the differences in geometric features in multiple dimensions such as position, angle, and length, it is determined whether the image is qualified. After the shooting is completed, the deviation of the container image can be immediately fed back. If the deviation exceeds the preset range, the system will generate a re-shoot reminder to notify the user to adjust the shooting angle or position. By analyzing the geometric features of the container outline and the shooting angle limiting frame, the shooting quality of each image can be ensured to be consistent. Each image undergoes the same quality inspection and deviation assessment to ensure the consistency of the image.

[0071] In some embodiments, the degree of deviation between each contour line and the corresponding limiting frame line is calculated according to the following formula (1):

[0072] (1)

[0073] in, For the The degree of deviation between the contour line and the corresponding line of the corresponding limiting frame line; For the contour lines; For the The limit frame line corresponding to the contour line; For the The directional similarity weight between the contour line and the corresponding limiting frame line; For the The position similarity weight of each contour line and the corresponding limiting frame line; For the The maximum distance between a contour line and the corresponding limit frame line;

[0074] No. The maximum distance between a contour line and the corresponding limit frame line is calculated according to the following formula (2):

[0075] (2)

[0076] in, For the The maximum distance between a contour line and the corresponding limit frame line; is the maximum value function; ; ; For the The endpoints of the contour lines To the corresponding limit box distance; For the The endpoints of the contour lines To the corresponding limit box distance; For the The contour lines correspond to the endpoints of the bounding box lines To Contour lines distance; For the The contour lines correspond to the endpoints of the bounding box lines To Contour lines distance.

[0077] Specifically, it is necessary to extract the geometric features of each contour line and the corresponding bounding box line from the image: contour line and bounding box The direction vector represents the direction of the line segment. The endpoint coordinates of the line segment are extracted to calculate the position difference between them. The direction similarity is calculated using cosine similarity. Specifically, cosine similarity can be calculated by dividing the dot product of the two direction vectors by their modulus. The maximum distance between the contour line and the bounding box line is calculated by To determine their positional deviations. Integrate the directional deviations and positional deviations to calculate the total deviation of each contour line from the corresponding limiting frame line. In the calculation of the maximum distance between the contour line and the corresponding limiting frame line: the i-th contour line is represented as ,in and are the coordinates of the two endpoints of the contour line. The i-th limit frame line is expressed as ,in and are the coordinates of the two endpoints of the limiting frame; the distance formula from the point to the straight line is used to calculate , , , These four distances are compared and the maximum value is taken to represent the farthest difference in geometric position between the two line segments. If the difference brought by the maximum value can meet the shooting requirements, it can be regarded that the position deviation meets the shooting requirements.

[0078] Through the method provided by this embodiment, by calculating the maximum distance between the endpoints of the contour line and the limiting frame line, the geometric difference between the two line segments can be accurately quantified. By calculating the directional similarity and positional similarity deviations, it is possible to automatically detect whether the contour of the container in each image meets the preset standards of the limiting frame, reducing the necessity of manual intervention. Improve work efficiency and avoid unnecessary subjective errors. By calculating the deviation between all contour lines and the limiting frame line, it is ensured that the quality of each captured image reaches a consistent high standard. This method reduces the problem of inconsistent quality caused by factors such as shooting angle and position.

[0079] In some embodiments, each upstream and downstream object type is parsed to determine the corresponding business needs; based on the business needs of each upstream and downstream object, information related to the needs of the corresponding upstream and downstream objects is filtered out from the coordinated data set; the filtered demand-related information is matched with the terminals of each object; based on different upstream and downstream object types, a suitable communication channel is determined and demand-related information is automatically sent.

[0080] Business requirements can be specific information required by upstream and downstream parties in the supply chain or business process, such as procurement requirements, inventory information, logistics status, order information, etc. Demand-related information can be data information that has been screened to match the specific business requirements of upstream and downstream parties. Communication channels can be the information transmission method or path used to transmit the screened demand-related information to the terminal devices of upstream and downstream parties.

[0081] Specifically, the different upstream and downstream object types in the business scenario are classified, such as suppliers, distributors, customers, etc. These types of classification help to determine the specific business needs of each object. Based on the type of each object, determine their specific business needs. For example, suppliers may need inventory and purchase order information, while customers may need the delivery status and invoice information of the order. Filter out information related to the business needs of upstream and downstream objects from the coordinated data set. The coordinated data set is a comprehensive database containing various types of information. Data that matches the object needs is extracted through preset filtering rules. After filtering out the information related to the needs of each upstream and downstream object, identify the terminal type of each object and match this information with the terminals of each object. For example, send an email to the supplier to update the order information, and send a text message to the customer to notify the order status. According to the different types of upstream and downstream objects, select the appropriate communication channel for information transmission. Communication channels can include email, text messages, application push instant messages, API docking, etc. After matching the terminals and determining the communication channels, automatically send the filtered demand-related information to the terminals of each upstream and downstream object through the appropriate communication channels.

[0082] Through the method provided by this embodiment, by accurately parsing the types and needs of upstream and downstream objects, it can be ensured that each object receives key information related to its business. This precise screening and matching avoids the problem of mistransmission or redundant transmission of information. It effectively improves the communication efficiency of each link in the upstream and downstream of the supply chain, and avoids misunderstandings or operation delays caused by inaccurate or untimely information. According to the preset rules, the system automatically completes the entire process from demand analysis, information screening, terminal matching to information sending, reducing the participation of manual operations. The operator does not need to check and send information one by one, which greatly simplifies the operation process. According to the needs and types of different upstream and downstream objects, the most appropriate communication channel is selected to ensure that the information is conveyed to the corresponding object in the best way. The flexibility and applicability of information transmission are improved, ensuring that the object can receive information in the most convenient and timely manner, and improving the reception rate and response speed of information. Through the system's automated information transmission mechanism, information delays or errors caused by human operations are reduced. For example, according to the screened needs and terminal matching rules, it is automatically ensured that each object obtains the required information in a timely manner.

[0083] In some embodiments, images in each set of images are classified to determine a number of information representative images and a number of status representative images; each information representative image is analyzed to determine the text area; each information representative image is cropped according to the text area and a first preset border range; the cropped image is deformed and corrected to obtain a number of corrected images; a number of similar corrected images in the same set of images are integrated to obtain an information card image; the free area in each status representative image is analyzed, and the corresponding information card image is embedded into the free area to form a combined image paste; the combined image paste is sorted and summarized to form an image collection.

[0084] The information representative image may be an image containing specific text, data or other information that needs to be extracted. The state representative image may be an image reflecting a certain state or scene of the container, such as intact, damaged or loaded state. The text area may be a part of the information representative image containing text or data information. The first preset frame range may be a boundary used to frame or limit the area of ​​interest in the image, which is used as a reference when intercepting and processing the image. Deformation correction may be an operation to correct the image distortion caused by shooting angle, light or other reasons, so that the image is restored to the correct proportion and shape. The corrected image may be an image that has been processed by deformity correction, and the distortion problems such as perspective and angle have been removed to present a standardized perspective. The information card image may be a processed and integrated image fragment, which usually contains important text or information content and is formed after correction and deduplication. The free area may be a part of the state representative image that is not occupied and can be embedded with other images or information. The combined image patch may be a final image formed by embedding the information card image into the free area of ​​the state representative image. The image set may be an ordered image data set integrated by multiple processed combined image patches.

[0085] Specifically, each set of images is analyzed using an image classification algorithm (such as an image classification model based on deep learning) for each image in each set of images; the images are classified into information representative images and status representative images according to the content features (such as text, shape, background, etc.) in the images. For each information representative image, the OCR (optical character recognition) technology is used to parse the information representative image and determine the text area therein. The coordinates and size of the text area are located according to the text area and the first preset border range, and then the image capture operation is performed on the identified text area to obtain the information representative image. The image processing algorithm is used to correct the deformity of the captured image to eliminate the image distortion caused by the shooting angle, perspective deformation, etc. The corrected images in the same set of images are analyzed for similarity. The image feature matching algorithm can be used to determine the similarity of the images. For the corrected images with a similarity higher than the preset threshold, they are merged to enhance the display effect and form an information card image. The free area in each state representative image is analyzed by the image segmentation algorithm. According to the size and shape of the free area, the information card image is scaled or adjusted to fit the embedding space, and the adjusted information card image is embedded into the free area of ​​the state representative image to form a combined image patch. All the generated combined image patches are sorted and summarized in a specific order or hierarchical structure to form a complete image set.

[0086] Through the method provided by this embodiment, by classifying and processing the information representative image and the state representative image, the image content can be better understood and analyzed, and the important information can be intercepted and processed according to the preset rules; the extraction efficiency and accuracy of the image information are improved, and unnecessary repeated operations and information loss are avoided. The image is automatically deformed to ensure that the image distortion problem caused by the shooting angle or other reasons is solved. The corrected image can be more standardized and easy to view. By integrating several similar corrected images in the same set of images, the system can effectively remove duplicate content and generate a single, centralized information card image. The system identifies the idle area in the state representative image and embeds the information card image therein, making full use of the idle space and avoiding the waste or stacking of information. This embedded integration method makes the image content richer and more organized. The blank area in the image is effectively utilized to optimize the visual effect of the image, improve the transmission efficiency of information, and save transmission bandwidth and storage space. By combining image stickers, multiple information is integrated into one image to form a more intuitive and integrated visual effect. Users can see related information in the same image, which improves the efficiency and effect of image display.

[0087] In some embodiments, each state representative image is parsed to determine the state representative image resolution, the main outline of the container in the image, and the extension direction; based on the extension direction, the free corners that are different from the extension direction are determined; based on the main outline and the free corners, the free area of ​​the largest triangle at the corners is determined; the information card image is parsed to obtain the information card image resolution and the minimum character size; based on the state representative image resolution, the minimum character display size of the state representative image is determined; the information card image resolution is adjusted based on the minimum character display size and the minimum character size; the information card image with adjusted resolution is embedded into the free area of ​​the largest triangle at the corners of the state representative image.

[0088] The resolution of the state representative image may be the pixel width and height of the state representative image. The main body outline may be the boundary or shape of the container in the state representative image. The extension direction may be the main extension direction of the state representative image caused by the perspective relationship. The free corner may be the area in the state representative image that is not occupied by the main body outline of the container. The resolution of the information card image may be the pixel width and height of the information card image. The minimum character size may be the size corresponding to the smallest character in the information card image. The minimum character display size may be the minimum size at which the characters in the state representative image can be clearly displayed.

[0089] Specifically, each state representative image is parsed to determine the resolution of the image; the image contour detection algorithm is used to parse the contour of the container or other subject in the image to determine its specific position and shape in the image. According to the shape and position of the subject, its extension direction in the image is analyzed; according to the extension direction of the subject contour, the idle corner area that is different from the extension direction is determined. This area is usually the part not occupied by the subject and is used as the idle area for embedding the information card image. If the extension direction of the subject is horizontal, the idle corners are usually located at the upper and lower corners of the image; if the extension direction is vertical, the idle corners are located at the left and right corners of the image. In most cases, the idle corners of the image are located in the upper left corner or the lower right corner due to the tilted extension caused by perspective. Therefore, according to the subject contour and the idle corners, the idle corner area is geometrically analyzed to determine the maximum triangle range of the idle area. This step can be performed by geometric calculation or by using the vertex coordinates of the corners to maximize the area. The resolution (width and height) of the information card image and the size of the smallest character in its content are obtained to ensure that the information card image can fit the idle area of ​​the state representative image to ensure the readability of the embedded information. According to the resolution of the state representative image, calculate its minimum character display size to ensure that the text in the information card image is clear and readable after embedding. According to the minimum character display size of the state representative image and the minimum character size of the information card image, use the image scaling algorithm to adjust the resolution of the information card image to adapt it to the size of the free area. Use image synthesis technology to embed the adjusted information card image into the free area of ​​the largest triangle at the corner of the state representative image to ensure that the information card image is visually harmoniously integrated with the state representative image.

[0090] Through the method provided by this embodiment, by analyzing the resolution, main body outline, extension direction and free area of ​​the state representative image, it is ensured that the information card image is accurately embedded in the best position of the state image. The whole process automatically identifies the available space in the image, making the embedding operation more intelligent and automated. No matter how the shape and size of the state representative image change, the best embedding area can be found, and the size of the information card image can be automatically adjusted according to the image resolution, achieving high adaptability. The adjusted information card image can be seamlessly integrated into the free area of ​​the state representative image, ensuring that the combined image is visually consistent and integrated naturally, avoiding the image from appearing abrupt or uncoordinated after embedding. By reasonably utilizing the triangular free area at the corners, the information card image will not interfere with the original main body outline or image structure, and the overall layout is more compact and beautiful. During the embedding process, by adjusting the resolution of the information card image and optimizing it according to the minimum character display size of the state representative image, it is ensured that the embedded information is clear and readable. The character size can be automatically adjusted to adapt to different resolutions to avoid blurry or illegible information, so that the embedded text and data content remain highly readable, and users can still clearly obtain information even in a smaller image area. By analyzing the main body outline and extension direction, the idle area is automatically identified, and the information card image is embedded using the largest triangle area at the corners.

[0091] In some embodiments, the cropped image is analyzed to determine the light and shadow structure of the cropped image; based on the light and shadow structure, the corrugated distribution structure of the surface of the container is determined; based on the corrugated distribution structure, the image is segmented, and the segmented parts are divided into a corrugated flat surface image and a corrugated inclined surface image, and are numbered according to the composition state; based on the position of the segmentation line between the corrugated flat surface image and the corrugated inclined surface image, the segmentation association points are determined; the corrugated flat surface image is perspective corrected to obtain a corrugated flat surface front view image; the projection relationship between the corrugated flat surface image and the corrugated inclined surface image is analyzed; the corrugated inclined surface image is corrected based on the corrugated flat surface front view image and the projection relationship to obtain a corrugated inclined surface front view image; the flat surface image and the corrugated inclined surface front view image are merged according to the numbering and segmentation association points to obtain a corrected image.

[0092] The light and shadow structure may be the distribution of light and dark areas formed on the surface of the container in the image due to different lighting. For example, on the surface of the container, the illumination makes the flat part of the corrugated brighter, while the inclined part is darker, forming an obvious light and shadow contrast. The corrugated distribution structure may be the distribution of corrugated protrusions and depressions on the surface of the container, including the relative position and arrangement pattern of the flat surface and the inclined part of the corrugated. The flat surface image of the corrugated may be an image of the flat part of the corrugated structure. The inclined surface image of the corrugated may be an image of the inclined part of the corrugated structure. The segmentation line position may be a boundary line between the flat surface of the corrugated and the inclined surface of the corrugated, which is used to segment the specific position of the flat surface image and the inclined surface image during image processing. The segmentation association point may be the intersection point of the flat surface image of the corrugated and the inclined surface image of the corrugated. . Perspective correction may be a specific operation of restoring the true form of an object in a frontal view by adjusting the perspective deformation in the image. The front view image of the flat surface of the corrugated may be an image of the flat surface of the corrugated restored to the frontal view after perspective correction. The projection relationship may be the relationship between the projection effects of the angle distortion caused by the bevel state between the flat surface of the corrugated board and the bevel surface. The front view image of the beveled corrugated board may be an image of the beveled corrugated board after projection analysis and perspective correction.

[0093] Specifically, during the photo-taking process, since the flat and inclined surfaces of the corrugated paper on the container surface, the gradient difference of light and shadow can reflect the convex and inclined parts of the corrugated paper, the image processing algorithm is first used to analyze the light and shadow structure of the selected image to identify the light and shadow distribution of different parts in the image. The flat and inclined surfaces of the corrugated paper on the container surface are identified by the change of the light and shadow boundary, and the convex and inclined parts of the corrugated paper are automatically identified according to the gradient difference of light and shadow. The specific position and boundary of each corrugated paper are located using the edge detection algorithm or the shape recognition algorithm. According to the distribution structure of the corrugated paper, the image is segmented and the image is divided into a flat corrugated surface image and a corrugated inclined surface image. Each segmented part is numbered so that the order and relevance of the image can be maintained in subsequent operations. On the boundary where the flat and inclined surfaces of the corrugated paper meet, the association points between the various corrugated areas are determined, and these points will be used for image correction and merging. The position of the association point is determined by calculating the intersection of the segmentation line or the end point of the boundary line. The flat corrugated surface image is corrected using a perspective transformation algorithm (such as a four-point perspective transformation, a matrix transformation, etc.). Adjust the four corner points of the image through the associated points to ensure that the boundary of the flat surface image becomes a straight line and restore the front view effect. Through the principle of geometric projection, analyze the angular distortion relationship between the flat surface and the inclined surface caused by the inclined surface state, and determine their projection relationship in the same plane. Calculate the degree of deformation of the inclined surface at different viewing angles, and provide data support for the subsequent correction of the inclined surface image. Use the perspective correction algorithm to adjust the perspective of the corrugated inclined surface image to eliminate the deformation of the inclined surface caused by the angle. According to the projection relationship between the corrugated flat surface and the inclined surface, adjust the angle and position of the inclined surface image to ensure consistency with the flat surface image. According to the previous numbering information, merge the corrugated flat surface image with the corresponding inclined surface image. Use the segmentation associated points to align the boundaries of the flat surface and the inclined surface image to ensure that the two are seamlessly connected to form a complete corrected image. For example: align the text that was originally on the same line.

[0094] Through the method provided by this embodiment, by perspective correction of the flat and inclined surface images of the corrugated board, the image deformation caused by the shooting angle or perspective problem is eliminated, so that the corrugated structure on the container surface is restored to the normal viewing angle. The image becomes more standardized and regular. By analyzing the light and shadow structure and the corrugated distribution structure, the system can accurately identify the dividing line between the flat and inclined surface of the corrugated board, and number and associate each part. In this way, when merging images, the precise alignment of different parts is ensured, and information loss or error is reduced. By correcting and merging the flat and inclined surface images of the corrugated board, the system seamlessly integrates the image parts of two different perspectives into a complete image. The corrected corrugated image has better continuity and consistency. The merged image is more complete, reducing the visual sense of discontinuity, especially when processing, displaying or archiving images, the overall effect is better. The corrected image ensures that the proportions and relationships between the various corrugated surfaces are more realistic, and the display effect is improved. Users can more intuitively understand the real situation of the container surface through the corrected image, avoiding misunderstandings caused by perspective distortion. Through precise segmentation and perspective correction, the details of the flat and inclined surfaces of the corrugated board are better preserved. Especially when merging images, the system can ensure the integrity and accuracy of the details based on the numbering and segmentation of associated points.

[0095] In some embodiments, the text range of each similar corrected image is identified; each similar corrected image is subjected to hue weakening and translucent processing to obtain an overlay base map; several overlay base maps are overlaid and integrated according to the text range, and the edges of the overlaid and integrated image are processed to obtain an easily recognizable image; the easily recognizable image is cropped according to a second preset border range to obtain an information card image.

[0096] The text range may be an area in an image that contains text content. Hue weakening may be an operation method of reducing the saturation of the color in the image, making the image look softer and reducing visual conflicts. Translucency may be an operation method of reducing the opacity of the image so that it can be better integrated when superimposed. The superimposed base map may be each similar image that has been processed with hue weakening and translucency to form an image base layer that can be superimposed. Superimposed integration may be combining several superimposed base maps according to specific rules to ensure that the text and important content of each layer of the image are clearly visible. Edge clear processing may be a processing method of enhancing the edges of the integrated image to make the outlines of each element in the image clearer, the details of the text and image more prominent, and avoiding the blurring effect caused by the superposition process. An easily recognizable image may be an image that has clear text and image content after superimposed integration and edge processing, and is easy for users to recognize and understand. The second preset border range may be an image cropping boundary set according to preset conditions, which may be stored in a preset database.

[0097] Specifically, due to the problem of shooting angle and light changes, some text in some images cannot be clearly identified during shooting, and the unrecognizable text in each image may be different. Therefore, OCR technology is used to scan the text part in the image and automatically identify the character area and boundary in the image. Mark the specific position and range of the text in each corrected image to ensure that the clarity of the text will not be disturbed during subsequent processing. Adjust the hue value of each similar corrected image to weaken the color intensity in each image to make it softer; make the image semi-transparent, usually by reducing the opacity of the image, so that multiple images can penetrate each other when superimposed. According to the recognition results of the text range, all base maps are superimposed according to a certain hierarchical structure to ensure that the text part of each layer will not be blocked. The superimposed image is optimized so that the text part and the image content can be naturally integrated while keeping each part clear and identifiable. Use the edge enhancement algorithm to process the edge of the superimposed image to make the image outline clearer. Focus on strengthening the edge of the text area to ensure that the text part is clearly visible and the image content is not blurred due to the superposition process. The cropping range of the image is determined according to the preset border size. The border range can be set according to the image content or design requirements, and is generally slightly smaller than the first preset border. The integrated image is accurately cropped using an image cropping algorithm.

[0098] Through the method provided in this embodiment, by weakening the hue and making it semi-transparent for each similar corrected image, the colors in the image become softer, and multiple images are smoother and more coordinated when superimposed, reducing visual conflicts and unnecessary contrasts. By identifying the text range and avoiding text occlusion when superimposed, it is ensured that the main information of each similar corrected image remains clearly visible after integration. Different image information is effectively arranged and adjusted to complement each other without overlapping. Through edge clarity processing, the text and edge details in the image are enhanced, making the integrated information card image easier to identify and understand, and the image details maintain high-quality display. Finally, a cropped and recognizable information card image is generated, ensuring the centralized display of the core content, and the main information content in the image remains prominent and clear.

Claims

1. A container information sharing system based on intelligent information technology, characterized in that: include: Acquire a first set of images of the loaded container; Performing image recognition processing on the first set of images to identify container number segment information on the container; Associating the container number segment information with the driver's work order information; Structuring container number segment information and related work order information to form a coordinated data set; acquiring a second set of images of the container prior to unloading; Performing embedded integration on the first set of images and the second set of images respectively to form an image set; Arrange the image set and the coordinated data set into a compilation archive; Send specific information in the coordinated data set to the object terminal according to the upstream and downstream object types; The acquisition of the first set of images of the container after loading and the second set of images of the container before unloading includes: Obtain the specific location of the container and the weather and ambient light at the corresponding location; According to the preset information collection requirements, the specific location of the container and the ambient light conditions, determine several shooting angle parameters of the mobile terminal and the shooting light parameters corresponding to each shooting angle; Determine a shooting angle limit frame according to shooting angle parameters; The shooting angle parameter, the shooting light parameter and the shooting angle limit frame are sent as shooting requirements to a mobile terminal; and an image of the container is captured by the mobile terminal; According to the shooting requirements, collecting a number of real-shot images of the mobile terminal; Analyze each of the real-shot images to determine whether the smear state in the real-shot image meets the standard; If the smear state cannot meet the standard, a reshoot request is generated and sent back to the mobile terminal; If the smear state meets the standard, the real-shot images are stored and used as the first set of images or the second set of images according to the shooting time node; The step of performing embedded integration on the first set of images and the second set of images to form an image set comprises: Classify the images in each set of images to determine a number of information representative images and a number of state representative images; Analyze each of the information representative images to determine the text area; Cutting and selecting each of the information representative images according to the text area and the first preset frame range; The deformity of the selected images is corrected to obtain a number of corrected images; Integrate several similar rectified images in the same set of images to obtain an information card image; Analyze the free area in each state representative image, and embed the corresponding information card image into the free area to form a combined image sticker; The combined image stickers are sorted and summarized to form an image collection.

2. The system according to claim 1, characterized in that The collecting, according to the shooting requirement, a plurality of real-shot images of the mobile terminal includes: Acquire the real-shot image from the mobile terminal; analyze the real-shot image to obtain the outline of the container in the image; Calculating the overall deviation between the container outline and the shooting angle limit frame; Determining whether the overall deviation degree is less than a preset deviation range; If the overall deviation is less than the preset deviation range, it is determined that the container outline meets the shooting angle limit frame in the shooting requirements; and the real shot images are collected; If the overall deviation is greater than the preset deviation range, it is determined that the container outline does not meet the shooting angle limit frame in the shooting requirements; a reshoot request is generated and sent back to the mobile terminal.

3. The system according to claim 2, characterized in that The calculating the overall deviation between the container outline and the shooting angle limit frame comprises: Analyzing the container outline and extracting the container outline into a plurality of contour lines; Analyzing the shooting angle limiting frame, and extracting the shooting angle limiting frame into limiting frame lines having the same number as the contour lines; Analyzing the similarity relationship between the container outline and the shooting angle limit frame; According to the similarity relationship, a plurality of contour lines are respectively bound to corresponding limiting frame lines; Calculate the degree of deviation between each contour line and the corresponding line of the corresponding limiting frame line; The overall deviation degree between the container outline and the shooting angle limit frame is determined according to the corresponding line deviation degree.

4. The system according to claim 3, characterized in that The calculation of the deviation degree between each contour line and the corresponding limiting frame line is calculated according to the following formula: ； in, For the The degree of deviation between the contour line and the corresponding line of the corresponding limiting frame line; For the contour lines; For the The limit frame line corresponding to the contour line; For the The directional similarity weight between the contour line and the corresponding limiting frame line; For the The position similarity weight of each contour line and the corresponding limiting frame line; For the The maximum distance between a contour line and the corresponding limit frame line; No. The maximum distance between a contour line and the corresponding limit frame line is calculated according to the following formula: ； in, For the The maximum distance between a contour line and the corresponding limit frame line; is the maximum value function; ; ; For the The endpoints of the contour lines To the corresponding limit box distance; For the The endpoints of the contour lines To the corresponding limit box distance; For the The contour lines correspond to the endpoints of the bounding box lines To Contour lines distance; For the The contour lines correspond to the endpoints of the bounding box lines To Contour lines distance.

5. The system according to claim 1, characterized in that The sending of specific information in the coordinated data set to the object terminal according to the upstream and downstream object types includes: Analyze each upstream and downstream object type to determine the corresponding business requirements; According to the business requirements of each upstream and downstream object, information related to the requirements of the corresponding upstream and downstream objects is filtered out from the overall data set; Match the filtered demand-related information with the terminals of each object; Determine the appropriate communication channel based on different upstream and downstream object types and automatically send demand-related information.

6. The system according to claim 1, characterized in that The step of analyzing the free area in each state representative image and embedding the corresponding information card image into the free area to form a combined image sticker includes: Analyze each of the state representative images to determine the resolution of the state representative image, the main outline of the container in the image, and the extension direction; According to the extension direction, determining the corners that are different from the extension direction as idle corners; According to the main body outline and the free corners, determine the free area of ​​the largest triangle at the corners; Parsing the information card image to obtain the information card image resolution and minimum character size; Determining the minimum character display size of the state representative image according to the state representative image resolution; adjusting the information card image resolution according to the minimum character display size and the minimum character size; The information card image with adjusted resolution is embedded into the free area of ​​the largest triangle at the corner of the state representative image.

7. The system according to claim 6, characterized in that The deformity correction of the selected image is performed to obtain a plurality of corrected images, including: Analyze the selected image to determine the light and shadow structure of the selected image; determine the corrugated distribution structure of the surface of the container according to the light and shadow structure; According to the corrugated distribution structure, the image is segmented, and the segmented parts are divided into a corrugated flat surface image and a corrugated inclined surface image, and are numbered according to the composition state; Determine a segmentation association point according to the position of a segmentation line between the corrugated flat surface image and the corrugated inclined surface image; Performing perspective correction on the corrugated flat surface image to obtain a front view image of the corrugated flat surface; Analyzing the projection relationship between the corrugated flat surface image and the corrugated inclined surface image; Correcting the corrugated oblique surface image according to the corrugated flat surface front view image and the projection relationship to obtain the corrugated oblique surface front view image; The flat surface image and the corrugated inclined surface front view image are merged according to the number and the segmentation associated points to obtain a corrected image.

8. The system according to claim 1, characterized in that The step of integrating a plurality of similar corrected images in the same set of images to obtain an information card image comprises: identifying a text range of each of said similar corrected images; Performing hue weakening and semi-transparency processing on each of the similar corrected images to obtain a superimposed base image; Overlay and integrate several overlay base maps according to the text range, and perform edge clear processing on the overlay and integrated image to obtain an easily recognizable image; The easily recognizable image is cropped according to the second preset frame range to obtain an information card image.

Images