A method for correcting sugar product barcodes using time-series analysis.

Abstract

This invention provides a highly efficient method for correcting sugar product barcodes by combining it with time-series analysis. [Solution] The method involves placing a scanner at station 0 to scan the product barcode, using a health scoring network to perform health scoring barcode correction on the scanned image in real time to create a product barcode score, obtaining the time-series ranking of the corresponding product, performing score discrimination on the product barcode score, and if the result of the score discrimination is lower than a predetermined score threshold, identifying the corresponding product as a weak barcode, triggering a station skip correction command to perform station barcode correction adaptive recognition, selecting a barcode correction station based on the adaptive optimization result, using that to retrieve product information based on the time-series ranking, and performing barcode correction processing on the corresponding product based on the result.

Description

[Technical Field]

[0001] The present invention relates to the field of automatic recognition, and more specifically to a method for correcting sugar product barcodes by combining time-series analysis. [Background technology]

[0002] Barcode technology is a crucial means of information identification in manufacturing and retail, and its reliability and accuracy significantly impact product tracking, inventory management, and production efficiency. In practical applications, barcode monitoring is widely used in automated production lines for sucrose and its derivative sugar products (white sugar, rock sugar, glucose, starch, etc.), where barcode data is read by scanning devices. However, due to limitations imposed by barcode print quality, scanning device performance, and the complexity of the production environment, recognition errors or data loss are likely to occur during barcode reading, affecting the integrity of data collection. To address this problem, research into automatic barcode correction technology is becoming an important direction for ensuring data integrity throughout the entire product process.

[0003] Conventional barcode correction methods typically rely on fixed rules or simple algorithms, limiting their ability to correct barcode information. In the initial stages of product barcode scanning, problems such as barcode smudges and unclear printing often cause scanners to fail to recognize or misidentify barcodes. Current solutions lack the ability to dynamically adjust correction policies and adequately integrate time-series information from the production line, resulting in data loss for some products and seriously impacting the accuracy and timeliness of subsequent product data processing. Therefore, a solution combining time-series analysis and intelligent correction mechanisms is needed to reduce barcode recognition error rates and improve the efficiency of production processes. [Overview of the Initiative] [Problems that the invention aims to solve]

[0004] This application aims to solve the technical problems of conventional technology, such as the unstable recognition quality of product barcodes and the inability to dynamically adjust correction policies, resulting in a high barcode error rate and low correction efficiency, by providing a method for correcting sugar product barcodes that combines time-series analysis. [Means for solving the problem]

[0005] In view of the above issues, this application provides a method for correcting sugar product barcodes by combining time-series analysis.

[0006] The sugar product barcode correction method combining time-series analysis according to the present application includes: placing a scanner at station 0 to scan the product barcode and create a real-time scan image; constructing a health scoring network, using the health scoring network to perform health scoring on the real-time scan image to create a product barcode score and obtain the time-series ranking of the corresponding product; performing score discrimination on the product barcode score, identifying the corresponding product as a weak barcode if the score discrimination result is lower than a predetermined score threshold, and triggering a station skip correction command; performing station barcode correction adaptive recognition based on the station skip correction command to create an adaptive optimization result; selecting a barcode correction station based on the adaptive optimization result, using the barcode correction station to retrieve product information based on the time-series ranking, and performing barcode correction processing for the corresponding product based on the result of the product information retrieval. [Effects of the Invention]

[0007] One or more technical means relating to this application have at least the following technical effects or advantages.

[0008] A sugar product barcode correction method is used that combines time-series analysis, which is a technical means that includes deploying scanners to create real-time scan images, performing scoring using a health scoring network, determining barcode quality and triggering station skip correction commands, and selecting barcode correction stations based on adaptive optimization results to perform barcode correction processing. This achieves the technical effects of improving barcode recognition accuracy, optimizing the correction process, and improving production efficiency.

[0009] The above description is merely an overview of the technical means of the present application. In order to understand the technical means of the present application more clearly, specific embodiments of the present application will be listed below, based on the contents of the specification, and to make the above and other objectives, features, and advantages of the present application easier to understand. [Brief explanation of the drawing]

[0010] [Figure 1] This is a schematic diagram showing the steps of a sugar product barcode correction method that combines time-series analysis according to an embodiment of the present invention. [Modes for carrying out the invention]

[0011] The overall concept of the technical means relating to this application is as follows:

[0012] An embodiment of the present invention provides a method for correcting sugar product barcodes by combining time-series analysis. Station 0 scans the barcode to create a real-time image, evaluates the quality of the barcode using a health scoring network, generates and identifies a score, marks barcodes below a predetermined threshold as weak barcodes, and triggers a station skip correction command. The optimal barcode correction station is selected through load and delay analysis of barcode correction stations, and product information is retrieved to complete the barcode correction operation.

[0013] After explaining the basic principle of the present application, various non-limiting embodiments of the present application will be specifically described below with reference to the drawings of the specification.

[0014] In an embodiment, as shown in FIG. 1, the method for correcting the barcode of a sugar product combined with time series analysis according to the embodiment of the present application includes the following steps S100 to S500.

[0015] In step S100, a scanner is arranged at station 0 to scan the product barcode and create a real-time scan image.

[0016] Specifically, station 0 refers to the first site on the production line and is usually used for initialization processes such as barcode scanning or product information input, and is the starting point of the production process. The scanner is a barcode recognition device that uses optical technology to scan and analyze the barcode into a digital signal and then transmit it to the data processing system. By arranging a high-precision barcode scanner at station 0 on the production line and configuring parameters such as the light source, focal length, and light-receiving element, the barcode images of each product passing through the production line can be quickly captured. Specifically, the scanner is fixedly mounted above the conveyor belt of station 0, and the distance between the scanner and the product surface, the scan angle, and the light source intensity are adjusted so that the scanner covers the optimal range of the product barcode area. When the product on the conveyor belt passes through station 0 and the sensor detects that the product enters the scan area, the scanner immediately triggers the shooting operation, collects the barcode image, and transmits it to the backend system in real time. For subsequent health scoring and recognition, operations including noise removal, contrast enhancement, trimming of the barcode area, and correction of the image tilt angle are performed on the barcode image transmitted from the scanner using the built-in algorithm.

[0017] By arranging a high-precision barcode scanner at Station 0 and creating barcode scan images in real time, the real-time state image of the barcode can be efficiently and accurately captured.

[0018] In step S200, a health scoring network is constructed, and the health scoring network is used to perform health scoring on the real-time scan image to create a product barcode score and obtain the chronological order of the corresponding product.

[0019] Specifically, the health scoring network is an artificial neural network based on deep learning. It is used to analyze the quality of the barcode image, evaluate the health status of the barcode, and provide a quantified scoring result. The product barcode score is a quantified indicator obtained based on the analysis result of the health scoring network. It is used to quantify the quality of the barcode image, and the higher the score, the better the health status of the barcode. The chronological order refers to the chronological arrangement position of the product on the production line and records the order in which the product passes through the scan site.

[0020] A health scoring model based on a convolutional neural network, such as a deep learning model built on TensorFlow or PyTorch, automatically extracts barcode features such as edge clarity, contrast, and continuity from real-time scanned images and outputs a score. The real-time scanned image is input to the health scoring network, and convolutional layers are used to extract important image features such as the smoothness, linearity, and continuity of the barcode lines. The network notes blurry areas of the barcode and calculates impact weights. Based on the extracted features, the health scoring network calculates an overall score using a weight model and stores the scoring results in a product database. Simultaneously, the system records the production time-series ranking of each product and establishes a one-to-one correspondence with the barcode score. The time-series position of products on the production line is updated in real time and used for subsequent station skip correction operations.

[0021] By implementing a health scoring network, the quality of barcode images can be accurately recognized, and a quantified scoring result can be generated for each product. Combined with the time-series ranking of the products, the system can quickly mark weak barcodes and plan a correction path. In step S300, a score determination is made to the product barcode score, and if the score determination result is lower than a predetermined score threshold, the corresponding product is identified as a weak barcode, and a station skip correction command is triggered.

[0022] Specifically, a weak barcode refers to a barcode whose score is lower than a predetermined score threshold, typically due to reasons such as blurred edges, breaks, discontinuities, or insufficient contrast, resulting in the barcode quality not meeting the standard. A station skip correction command refers to a command generated by the system after a barcode has been marked as a weak barcode, instructing the production line to skip the current station and move to a designated barcode correction station to perform a barcode correction operation.

[0023] After obtaining the product barcode score, it is compared to a predetermined score threshold. If the score is below the threshold, it is determined to be a weak barcode; if the score is above the threshold, it is determined to be a normal barcode and no correction is necessary. For products determined to have weak barcodes, a station skip correction command is generated for the products marked as having weak barcodes, instructing the production line to skip the current station and move to a designated barcode correction station.

[0024] By using score determination, weak barcode identification, and triggering station skip correction commands, it is possible to quickly sort barcodes that need correction and accurately guide them to the appropriate barcode correction station, thereby improving the level of automation and barcode correction efficiency of the production line.

[0025] In step S400, station barcode correction adaptive recognition is performed based on the above station skip correction command, and an adaptive optimization result is created.

[0026] Specifically, adaptive station barcode correction recognition refers to dynamically evaluating each station on the production line based on its actual state to select the optimal station for barcode correction. Adaptive optimization results refer to calculating the optimal station by combining station load analysis and delay constraints to ensure that the barcode correction operation strikes the best balance between production efficiency and accuracy. Delay time refers to the additional time required for product barcode correction, including waiting due to queuing, distribution delay time due to transportation between stations, and queue delay time. Constraints refer to the maximum allowable range of set load and delay time, and are used to select available barcode correction stations.

[0027] Upon receiving a station skip correction command, the system recognizes the barcode product requiring barcode correction and its current station, and retrieves the barcode correction station information database. Load analysis and delay time calculations are performed sequentially for all potential barcode correction stations on the production line. Stations that meet the constraints are selected based on the maximum load limit and delay time limit of the production line. From the stations that meet the conditions, the station with the lightest load and shortest delay time is selected as the barcode correction station. The optimization results, including the optimal station number, estimated delay time, and barcode correction operation time, are output.

[0028] By combining station skip correction commands with station barcode correction adaptive recognition, the barcode correction policy can be dynamically adjusted, ensuring the efficiency and stability of the production line. This avoids resource waste and production delays caused by random station selection.

[0029] In step S500, a barcode correction station is selected based on the adaptive optimization results, product information is retrieved based on chronological order using the barcode correction station, and barcode correction processing is performed on the corresponding product based on the results of the product information retrieval.

[0030] Specifically, the time-series order refers to the arrangement order of products on the production line and is used to track the production status and distribution information of each product, and to ensure correspondence between product information retrieval and barcode correction processing. Product information retrieval refers to extracting product-related information, including product number, barcode image, and its health score, from the database based on the time-series order and is used to guide barcode correction correction. Barcode correction processing refers to performing correction or regeneration operations on the barcode at the barcode correction station based on the results of product information retrieval, such as reprinting the barcode, correcting errors, or supplementing information.

[0031] Based on the adaptive optimization results, a target barcode correction station is determined, and a conveyor belt or robotic arm is controlled to transport the product to that station. During the transport process, the time-series ranking of the products is updated synchronously to ensure that the barcode correction operation matches the product information. After the product arrives at the barcode correction station, specific information related to the product, including the product number, barcode image, health score, and recognized barcode defect type (e.g., blurry, discontinuous), is extracted from the database based on the time-series ranking. Based on the retrieved product information, the barcode correction station performs the following processes on the barcode using a predetermined correction mechanism: regenerating a clear barcode using a barcode printer and reprinting the barcode to cover the old barcode; optical error correction to correct blurry or broken barcode lines by laser engraving; and rescanning to re-recognize barcodes that failed to be read by optimizing the scan angle or adjusting the light source. Specifically, a high-precision barcode printer, laser engraver, or high-resolution scanner can be installed in the barcode correction station, and efficient barcode correction can be achieved in combination with image processing software. After the barcode correction process is complete, the barcode correction operation results, including information such as the barcode status before and after correction and the time taken for correction, are recorded for subsequent tracking and optimization processes.

[0032] By selecting a barcode correction station based on the adaptive optimization results and retrieving product information to perform barcode correction processing, the accuracy and efficiency of barcode correction operations can be ensured, and manual intervention can be reduced. Furthermore, creating a product barcode score by performing health scoring of the real-time scanned image using the health scoring network includes: positioning the region of interest of the real-time scanned image using the positioning preprocessing layer of the health scoring network and creating a region of interest positioning result; performing edge contour extraction on the region of interest positioning result and performing edge smoothness scoring based on the edge contour extraction result to generate a first scoring result; performing edge linearity scoring based on the edge contour extraction result to create a second scoring result; performing edge continuity scoring based on the edge contour extraction result to create a third scoring result; performing contrast recognition between the region of interest positioning result and the background on the real-time scanned image to create a fourth scoring result; and creating a product barcode score based on the first scoring result, the second scoring result, the third scoring result, and the fourth scoring result.

[0033] Specifically, the positioning preprocessing layer refers to the first layer in the health scoring network, which is used to improve the accuracy and efficiency of subsequent feature analysis by recognizing the region of interest in the barcode image and removing background parts unrelated to the barcode. Edge smoothness scoring refers to calculating the smoothness of the barcode based on the angle changes of the barcode lines, with a high score indicating that the barcode lines are smooth and do not change abruptly. Edge linearity scoring evaluates the linearity characteristics of the barcode lines and quantifies whether the barcode has good linearity using the coefficient of determination of linear fitting. The edge continuity score is used to evaluate whether there are breaks or gaps in the barcode lines, with a high score indicating that there are no obvious breaks in the barcode lines.

[0034] After a barcode image is input to the health scoring network, the positioning preprocessing layer first recognizes the region of interest of the barcode. This process is achieved by combining sliding window and convolutional operations. For example, the barcode region is separated from the background using the OpenCV image segmentation algorithm. The Canny edge detection method extracts the barcode's edge lines and generates a contour image. The results of contour extraction provide a basis for subsequent scoring. For example, for a blurry barcode, contour extraction recognizes most of the lines, but gaps and blurry areas still exist. The contour lines are analyzed, the angular change between adjacent points on the line is calculated, and a score is assigned based on the stability of the change. Barcodes with small angular changes receive higher scores. Linear regression is used to fit the contour lines, and the coefficient of determination is calculated to quantify linearity. Higher fitting accuracy results in higher scores. The number of gaps and the maximum gap length in the contour are detected, and the gap status is quantified with a score. Barcodes with fewer gaps and good continuity receive higher scores.

[0035] By analyzing the difference in grayscale values ​​between the barcode area and the background, contrast is quantified and scored. For example, the difference in the average grayscale values ​​between the barcode area and the background can be calculated. Smoothness, linearity, continuity, and contrast scores are combined, and a barcode score is generated based on weighted ratios. By using a health scoring network to perform multidimensional analysis on real-time scanned images, barcode quality can be accurately and comprehensively evaluated, a scientific score can be generated, and a basis for subsequent corrective actions can be provided. Furthermore, the above first scoring result is calculated using the following formula:

number

number

[0036] Specifically, the edge smoothness score is a score used to evaluate the smoothness of the barcode's edge lines; a higher score indicates smoother edges, more uniform lines, and better quality. The average angle change refers to the average value of the angle change between adjacent lines on the barcode's edge and reflects the smoothness of the line. A small angle change indicates smooth line connections. The edge angle change threshold is a reference value for evaluating smoothness, and is usually a predetermined standard that indicates the range within which a line can be considered smooth. The direction angle refers to the direction from one point on the edge line to the other and is used to describe the trend of the barcode line. The total number of points refers to the number of all points on the edge line and indicates the length and complexity of the barcode line.

[0037] First, the coordinates (x) of all points on the edge line of the barcode i , y i By extracting the direction angle θ of adjacent points in each pair, i To calculate, specifically,

number

[0038] Edge smoothness scoring can quantify the quality of barcode lines. In particular, when the barcode is broken or not smooth due to printing blurriness or damage, the score will clearly decrease. Such a scoring mechanism can effectively identify the barcodes that need to be corrected, thus improving the accuracy of barcode scanning.

[0039] Furthermore, the above second scoring result is calculated by the following formula:

Number

Number

Number

Number

Number

[0040] Specifically, the edge linearity score is used to evaluate whether the edge line of the barcode exhibits good linear characteristics. The higher the score, the closer the barcode line is to a straight line, indicating better quality. The coefficient of determination is a statistical quantity for evaluating the degree of fitting between the fitting line and the actual edge points. The closer the value is to 1, the closer the distribution of the edge points is to a straight line, indicating a high fitting effect. The minimum criterion of the coefficient of determination is a threshold for evaluating whether the fitting result meets the minimum requirement of linearity.

[0041] Using the least squares method, a linear fitting is performed on the extracted edge points to generate the equation of the fitting line, and the actual vertical coordinate y of the edge points is obtained. i and the predicted values ​​of the fitting line

number

number

[0042] Edge linearity scoring effectively identifies problems with line curvature due to reduced print quality or damage by evaluating whether the overall tendency of the barcode edges is close to straight. Furthermore, the above third scoring result is calculated using the following formula:

number

number

[0043] Specifically, the edge continuity score is used to evaluate whether the edge lines of a barcode are continuous and free from obvious breaks or gaps. A higher score indicates greater edge continuity and higher barcode quality. The maximum gap length refers to the maximum distance between adjacent points on the edge line of a barcode and indicates the severity of the edge break or gap. A smaller value indicates a smaller gap and greater edge continuity. The threshold for the acceptable maximum gap length is a predetermined standard and is used to determine whether there are unacceptable gaps in the edge. gap If this threshold is exceeded, it indicates that the barcode quality is poor.

[0044] The distance between adjacent points in each pair is calculated sequentially, and the distance is calculated using the Pythagorean theorem based on the coordinates of the two points. The resulting intervals between all points are recorded and used to find the maximum distance (i.e., the maximum gap length). The maximum gap length G for all point pairs is calculated. gap Find the threshold G for the maximum allowable gap length. threshold Compare with G gap If the gap exceeds the threshold, it indicates that the continuity of the barcode edges is low and that a correction process needs to be triggered. A score is calculated based on the ratio of the gap length to the threshold. The score ranges from 0 to 100, with a higher score indicating a smaller maximum gap.

[0045] By calculating the edge continuity score, the integrity of barcode edges can be accurately assessed, and breaks and gaps can be automatically detected. Combined with the scoring results, it is possible to quickly determine whether or not the barcode needs to be corrected.

[0046] Furthermore, performing score determination on the above-mentioned product barcode score includes generating a normal transmission command if the score determination result satisfies a predetermined score threshold, and executing the transmission process for the corresponding product based on the normal transmission command.

[0047] Specifically, after receiving the barcode score, it is compared to a predetermined score threshold. If the score is above the threshold, the barcode is deemed acceptable; if it is below the threshold, it is marked as a weak barcode that requires correction. For barcodes determined to be acceptable, a normal transmission command is automatically generated and transmitted to the control module of the conveyor belt or production equipment via an industrial communication protocol. This command may include the product number, barcode score, and target site information. After receiving the normal transmission command, the production equipment starts the corresponding operation and smoothly transmits the product to the next site. The operating status of the equipment is monitored in real time by sensors to ensure that there are no errors in the transmission process. While executing the transmission process, the score and transmission command are recorded in a database for subsequent analysis and quality tracking.

[0048] By performing score discrimination against scores and generating and executing normal transmission commands, rapid sorting and efficient processing of barcode quality can be achieved, significantly improving the automation level of the production line and the accuracy of barcode transmission.

[0049] Furthermore, creating adaptive optimization results by performing station barcode correction adaptive recognition based on the above station skip correction command includes: performing barcode correction load analysis on the station, including continuous load analysis and temporary window load analysis, to create load constraints; obtaining the delay time of station barcode correction, including distribution delay time and barcode correction queue delay time, to create delay constraints; and performing station barcode correction adaptive optimization based on the above load constraints and delay constraints to generate adaptive optimization results.

[0050] Specifically, continuous load analysis refers to evaluating the average operating load of a station under stable operating conditions. Temporary window load analysis refers to analyzing the additional load increased at a station by an unexpected task within a specific period. Distribution delay time is the time required to transport a product from its current station to the barcode correction station, and is determined by the conveyor belt speed and transport distance. Barcode correction queue delay time refers to the delay time that occurs when a product is waiting at the barcode correction station due to task queuing. Load constraints refer to load limits set based on the operational capacity of the production line; if a station's current load exceeds this threshold, it is considered to not meet the requirements. Delay constraints refer to set delay time limits; if a station's distribution delay time or queue delay time exceeds this constraint, the station is removed.

[0051] Upon receiving a barcode correction command, the system analyzes information about products requiring barcode correction via station skipping and the stations where they are located, initiating a station status evaluation process. A real-time data acquisition module evaluates the current operational load of all candidate barcode correction stations, i.e., calculates the average workload currently being handled by each station. Distribution delay time is calculated by collecting conveyor belt speed and transport distance using sensors, and barcode correction queue delay time is calculated based on the quantity of queued products and processing speed. Inter-station distribution information is collected in real time by the PLC controller and sensors. The load and delay time of all candidate stations are compared with predetermined constraints, and stations exceeding the limits are eliminated. From the stations that meet the conditions, the optimal barcode correction station is selected based on the principles of load minimization and delay time minimization, and specific barcode correction means are generated. The adaptive optimization results are transmitted to the production equipment, and the conveyor belt or robot arm is controlled to send products to the optimal barcode correction station, while monitoring the barcode correction operation in real time.

[0052] By combining station skip correction commands with station barcode correction adaptive recognition, intelligent scheduling and optimization of production line resources are achieved. The optimal station can be dynamically selected to perform barcode correction, significantly reducing production line delays and resource waste.

[0053] Based on the above, the sugar product barcode correction method incorporating time-series analysis according to the embodiment of the present application significantly improves the accuracy of barcode recognition and the completeness of production data by dynamically adjusting the scanning, scoring, and correction policies. By monitoring the health status of barcodes in real time, it ensures that quality problems can be detected and addressed in a timely manner, reduces the barcode misjudgment rate, optimizes production efficiency, and is particularly suitable for automated production lines where high precision is required.

[0054] The above description is merely a preferred embodiment of the present invention and does not limit the present invention in any way. The present invention has been disclosed as described above by preferred embodiments, but does not limit the present invention. Those skilled in the art can create equivalent embodiments by modifying or changing the disclosed technical content and making equivalent substitutions without departing from the scope of the technical means of the present invention. Any modifications, equivalent substitutions, and alterations made to the above embodiments in accordance with the technical idea of ​​the present invention, as long as they do not depart from the technical means of the present invention, should all be included within the scope of the technical means of the present invention. [Explanation of Symbols]

[0055] Queue delay time Scoring results barcode image Scan area Delay time

Claims

1. The scanner is placed at Station 0 to scan the product barcode and create a real-time scan image, The system involves constructing a health scoring network, using the health scoring network to perform health scoring on the real-time scanned images, creating product barcode scores, and obtaining the corresponding time-series ranking of the products. The system performs score discrimination on the aforementioned product barcode score, and if the result of the score discrimination is lower than a predetermined score threshold, it identifies the corresponding product as a weak barcode and triggers a station skip correction command. Based on the aforementioned station skip correction command, the system performs station barcode correction adaptive recognition to create an adaptive optimization result. This includes selecting a barcode correction station based on the adaptive optimization results, using the barcode correction station to retrieve product information based on chronological order, and performing barcode correction processing for the corresponding product based on the results of the product information retrieval. Using the aforementioned health scoring network to perform health scoring on the real-time scanned images and create a product barcode score is: The positioning preprocessing layer of the health scoring network is used to position the region of interest in the real-time scan image and to create a region of interest positioning result. The process involves performing edge contour extraction on the aforementioned region of interest positioning result, performing edge smoothness scoring based on the edge contour extraction result, and generating a first scoring result. Based on the aforementioned edge contour extraction results, edge linearity scoring is performed to create a second scoring result. Based on the edge contour extraction results, edge continuity scoring is performed to create a third scoring result. The real-time scan image is subjected to contrast recognition between the region of interest positioning result and the background, and a fourth scoring result is created. A method for correcting sugar product barcodes, combined with time-series analysis, characterized by comprising creating a product barcode score based on the first scoring result, the second scoring result, the third scoring result, and the fourth scoring result.

2. The first scoring result is calculated using the following formula: [Math 1] [Math 2] In the formula, S ｓｍｏｏｔｈｎｅｓｓ This represents the edge smoothness score, and Δθ ｍｅａｎ Δθ represents the average angle change between edge points. ｔｈｒｅｓｈｏｌｄ θ represents the edge angle change threshold, n represents the total number of points on the same line, i represents any point on the current line, and ｉ θ represents the direction angle from point i to point i+1. （ｉ－１） The method for correcting sugar product barcodes, which combines time-series analysis according to claim 1, is characterized in that represents the direction angle from the (i-1)th point to the (i)th point.

3. The second scoring result is calculated using the following formula: [Math 3] [Math 4] In the formula, S ｌｉｎｅａｒｉｔｙ This represents the edge linearity score, R ２ This represents the coefficient of determination of the fitting line for the edge points, [Math 5] This represents the minimum standard for the coefficient of determination, [Math 6] This represents the predicted value on the fitting line, [Number 7] This represents the average y value of all points, and y ｉ The method for correcting sugar product barcodes, which combines time-series analysis as described in claim 1, is characterized in that represents the actual vertical coordinate value of the i-th point.

4. The third scoring result is calculated using the following formula: [Number 8] [Number 9] In the formula, S ｃｏｎｔｉｎｕｉｔｙ represents the edge continuity score, G ｇａｐ represents the maximum gap length of the edge contour, G ｔｈｒｅｓｈｏｌｄ represents the threshold value of the maximum allowable gap length, x ｉ＋１ represents the actual abscissa value of the (i + 1)-th point, xi represents the actual abscissa value of the i-th point, y ｉ＋１ is the ｉ＋１ actual ordinate value of the point, and a method for correcting a sugar product barcode by combining the time series analysis according to claim 1, characterized in that.

5. Performing score discrimination on the aforementioned product barcode score means If the score determination results in a predetermined score threshold being met, a normal transmission command is generated. A method for correcting sugar product barcodes, which combines time-series analysis according to claim 1, comprising performing a transmission process for the corresponding product based on the aforementioned normal transmission command.

6. Performing station barcode correction adaptive recognition based on the aforementioned station skip correction command and creating an adaptive optimization result is performed. Perform barcode-corrected load analysis on the station, including continuous load analysis and temporary window load analysis, and create load constraints. Obtain the delay time for station barcode correction, including the distribution delay time and the barcode correction queue delay time, and create delay constraints. A method for correcting sugar product barcodes by combining time series analysis according to claim 1, characterized in that it includes performing adaptive optimization of station barcode correction based on the load constraint and the delay constraint to generate adaptive optimization results.

Images