Seamless steel tube continuous rolling production process tracing method

Abstract

The invention belongs to the technical field of production process tracing, and particularly relates to a seamless steel tube continuous rolling production process tracing method, which comprises the following steps of: aiming at a target node in a seamless steel tube continuous rolling production process, acquiring a product image of a semi-finished product in the target node, and according to the target node and a secret key of an adjacent production node, acquiring a product image of the semi-finished product in the target node; the method comprises the steps of obtaining first secret information from a concern area of a product image, embedding the first secret information into a marking area through a steganography algorithm to obtain a secret image, and when a production process is traced, obtaining second secret information from a concern area of the secret image according to a traced node and a secret key of an adjacent production node, and a verification sequence is extracted from the marked area of the secret image through a steganography algorithm, and the product image of the traced node is stored completely and truly in response to the fact that the verification sequence is the same as the second secret information. According to the invention, the accuracy of determining the responsibility attribution by tracing according to the product image in the production process is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of production process tracing, and more particularly to a method for tracing the production process of a seamless steel pipe continuous rolling mill. Background Art

[0002] The seamless steel pipe continuous rolling production line is a highly automated and integrated production system. Each production node is closely connected to achieve high-efficiency and high-quality steel pipe production. Between different production nodes, due to production deviations, equipment failures, operational errors and other factors, the products produced may have defects in subsequent inspection and use.

[0003] In the related technology, for example, the Chinese patent application document with application publication number CN118691167A discloses a quality traceability management system and method for the entire production process of aluminum profiles, including a data comparison unit, a data query unit and a quality traceability unit, and also includes: a node definition unit, the node definition unit is used to define nodes for the entire production process of aluminum profiles to obtain a production node set for the entire production process of aluminum profiles, the production node set includes raw material procurement nodes, smelting workshop nodes, extrusion molding workshop nodes, heat treatment workshop nodes, surface treatment workshop nodes, quality inspection nodes and transportation nodes, and the production node set of the entire production process of aluminum profiles is transmitted to the blockchain construction unit and the real-time monitoring unit.

[0004] In related technologies, when a product has quality problems, responsibility is determined based on product images during the production process. However, if the product images are tampered with, the relevant production nodes can evade accountability. To improve the accuracy of accountability, it is necessary to ensure that the product images of each production node are preserved intact and authentic. Summary of the Invention

[0005] In order to solve the above-mentioned technical problem that when product quality problems occur, if the product image in the production process is tampered with, the relevant production nodes can escape accountability, the present invention provides a seamless steel pipe continuous rolling production process tracing method, comprising: for a target node among multiple production nodes in the seamless steel pipe continuous rolling production process, collecting the product image of the semi-finished product in the target node, the product image comprising multiple focus areas and one annotation area; obtaining first secret information from the focus area of ​​the product image according to the key of the target node and the production nodes adjacent thereto, comprising: obtaining segmentation thresholds in each direction according to the distribution of the characteristic values ​​of all historical pixel points in each direction in the focus area of ​​the historical image of the target node; for the pixel points in the focus area, obtaining the segmentation thresholds in each direction according to the size relationship between the characteristic values ​​of the pixel points in each direction and the segmentation thresholds in each direction, A characteristic identifier of a pixel is obtained, where the characteristic value of the pixel in each direction refers to the difference between the grayscale value of the pixel and the grayscale values ​​of other pixels in the neighborhood of the pixel in each direction, and the characteristic identifier is a binary number; a bit operation is performed on the characteristic identifiers of all pixels according to the keys of the target node and the production nodes adjacent to it to obtain a first secret information; the first secret information is embedded in the marked area through a steganographic algorithm to obtain a secret image; when tracing the production process, a bit operation is performed on the characteristic identifiers of all pixels in the focus area of ​​the secret image according to the keys of the traced node and the production nodes adjacent to it to obtain a second secret information; a verification sequence is extracted from the marked area of ​​the secret image through a steganographic algorithm; in response to the verification sequence being the same as the second secret information, the product image of the traced node is preserved intact and authentic.

[0006] In the present invention, the focus area of ​​the product image is the main area in the product image of the semi-finished product processed by the previous production node that affects the product quality of the semi-finished product of the next production node. Therefore, the first secret information is obtained from the focus area of ​​the product image and embedded in the marked area of ​​the product image. When the production process is traced, when the focus area that affects the product quality is tampered with, the second secret information obtained from the focus area of ​​the secret image is different from the verification sequence extracted from the marked area of ​​the secret image, thereby ensuring that the product image of each production node is preserved intact and authentic. When quality problems occur in subsequent products, the accuracy of determining the responsibility based on the product image during the production process is improved. Among them, since the product image is associated with two adjacent production nodes and is an important basis for dividing the responsibility between the two adjacent production nodes, the present invention obtains the first secret information from the focus area of ​​the product image according to the key of the target node and the production node adjacent to it, which can ensure that no matter which node of the two adjacent production nodes wants to tamper with the product image, it will be discovered, thereby ensuring that the product image of each production node is preserved intact and authentic.

[0007] Preferably, the multiple production nodes in the seamless steel pipe continuous rolling production process include raw material preparation nodes, raw material heating nodes, raw material processing nodes, piercing nodes, rolling nodes, sizing nodes, cooling nodes, straightening nodes, cutting nodes, and surface treatment nodes.

[0008] Preferably, the semi-finished product in the target node refers to the semi-finished product processed by the previous production node of the target node; the product image of the semi-finished product in the target node includes the product image of the semi-finished product in the target node before being processed by the target node and the product image of the semi-finished product in the target node after being processed by the target node; the area of ​​interest included in the product image before processing is specified by the target node, and the area of ​​interest included in the product image after processing is specified by the next production node of the target node.

[0009] The focus area of ​​the product image in the present invention is the main area in the product image of the semi-finished product after processing by the previous production node that affects the product quality of the semi-finished product of the next production node. Therefore, the focus area in the product image before processing in the target node is specified by the target node, which can ensure that the main area affecting the product quality of the semi-finished product of the target node is focused on, thereby ensuring that when the target node is traced, the product image of the target node is preserved intact and authentic, and can more accurately determine whether it is the responsibility of the target node, thereby improving the accuracy of determining the responsibility by tracing the product image during the production process.

[0010] Preferably, the key of the production node is set according to the initial conditions of the two-dimensional chaotic mapping function, including: evenly dividing the parameters and initial values ​​in the initial conditions of the two-dimensional chaotic mapping function into two groups; for any production node, if the serial number of the production node is odd, randomly generating corresponding parameters and initial values ​​from the value range of the parameters and initial values ​​of the first group as the key of the production node; if the serial number of the production node is even, randomly generating corresponding parameters and initial values ​​from the value range of the parameters and initial values ​​of the second group as the key of the production node.

[0011] Preferably, the step of obtaining the segmentation thresholds in each direction includes: obtaining all historical images of the target node; Direction to In any direction, the characteristic values ​​of all historical pixels in the focus area of ​​the historical image in that direction are obtained; the range of the separation value is , The minimum and maximum values ​​of the characteristic values ​​of all historical pixel points in this direction are respectively used. According to the size relationship between the characteristic values ​​of the historical pixel points in this direction and the separation value, all historical pixel points are divided into two categories, including: all historical pixel points with characteristic values ​​in this direction less than the separation value are divided into the first category, and all historical pixel points with characteristic values ​​in this direction greater than or equal to the separation value are divided into the second category; the preference of different separation values ​​is calculated, and the separation value with the largest preference is used as the segmentation threshold for this direction.

[0012] Preferably, the preference of the different separation values ​​satisfies the expression: Where, Separator value The preference of are the intra-class variances of the eigenvalues ​​of all historical pixels in the first and second categories in this direction, is the inter-class variance of the eigenvalues ​​of all historical pixels in the first and second categories in this direction, is the natural exponential function, is the mean of the characteristic values ​​of all historical pixels in the category with the least number of historical pixels in this direction, To take the absolute value.

[0013] In the present invention, all historical pixel points are divided into two categories according to the size relationship between the characteristic value of the historical pixel point in this direction and the separation value. Among the two categories obtained, the number of categories representing products with poor quality is the least. When calculating the preference of different separation values, when the separation value is The smaller the difference between the mean of the characteristic values ​​of all historical pixels in the category with the least number of historical pixels in this direction, the smaller the separation value. The preference The larger the value, the closer the final selected segmentation threshold is to the characteristic value of the poor quality product. This ensures that when you want to tamper with the product image of the poor quality product in the future, the characteristic identification of the pixel point is more easily destroyed, making the tampering easier to detect.

[0014] Preferably, the first The bits satisfy the expression: Where, is the first feature identifier of the pixel point bits, For this pixel point The eigenvalue of the direction, For the The segmentation threshold of the direction, for Integers in the range, is the preset second value.

[0015] The present invention obtains the characteristic identification of the pixel points based on the local characteristics of the pixel points in the area of ​​interest. When the grayscale value of the pixel points in the area of ​​interest is tampered with, the local characteristics of the relevant pixel points will change, which in turn causes the characteristic identification of the relevant pixel points to change. This ensures that the product images of each production node are preserved intact and authentic. When quality problems occur in subsequent products, the accuracy of determining the attribution of responsibility through tracing back the product images during the production process is improved.

[0016] Preferably, the characteristic values ​​of the pixel points in each direction include: for any pixel point in the area of ​​interest, the characteristic values ​​of the pixel point with the pixel point as the center and the size equal to The neighborhood of the pixel is evenly divided to obtain the pixel point in Direction to Neighborhood of direction, The first and second values ​​are preset respectively; the pixel point is Eigenvalue of direction Satisfies the expression: Where, is the grayscale value of the pixel, For this pixel point The mean of the grayscale values ​​of all other pixels in the neighborhood of the direction, To take the absolute value, for Integers in the range, is the preset second value.

[0017] Preferably, the obtaining of the first secret information comprises: using the secret keys of the target node and the production nodes adjacent thereto as initial conditions of a two-dimensional chaotic mapping function, iterating the two-dimensional chaotic mapping function times, and each iteration obtains a chaotic value pair. Respectively The first chaotic value and the second chaotic value in the chaotic value pair; obtaining the first secret information requires Secondary operation, is equal to the number of all pixels in the area of ​​interest. The result of the second operation is used as the first secret information; When performing the secondary operation, , will The result of the secondary operation Bit and The feature identifier of the pixel The bits are aligned and then bit operations are performed; , , For the The number of all bits in the result of the bitwise operation and the number of all bits in the feature identifiers of all pixels are equal to , is the preset second value.

[0018] The present invention uses the keys of two adjacent production nodes associated with the product image to perform bit operations on the feature identifiers of all pixels in the focus area to obtain the first secret information. This can ensure that no matter which of the two adjacent production nodes wants to tamper with the product image, it will be discovered, thereby ensuring that the product image of each production node is preserved intact and authentic.

[0019] Preferably, the method also includes: for the target node, the production nodes adjacent to it include the previous production node of the target node and the next production node of the target node. Therefore, for the product image in the target node before processing, the key based on the target node and the production nodes adjacent to it refers to the key based on the target node and the key of the previous production node of the target node; for the product image in the target node after processing, the key based on the target node and the production nodes adjacent to it refers to the key based on the target node and the key of the next production node of the target node.

[0020] The beneficial effects of the present invention are: In the present invention, the focus area of ​​the product image is the main area in the product image of the semi-finished product processed by the previous production node that affects the product quality of the semi-finished product of the next production node. Therefore, the first secret information is obtained from the focus area of ​​the product image and embedded in the marked area of ​​the product image. When the production process is traced, when the focus area that affects the product quality is tampered with, the second secret information obtained from the focus area of ​​the secret image is different from the verification sequence extracted from the marked area of ​​the secret image, thereby ensuring that the product image of each production node is preserved intact and authentic. When quality problems occur in subsequent products, the accuracy of determining the responsibility based on the product image during the production process is improved. Among them, since the product image is associated with two adjacent production nodes and is an important basis for dividing the responsibility between the two adjacent production nodes, the present invention obtains the first secret information from the focus area of ​​the product image according to the key of the target node and the production node adjacent to it, which can ensure that no matter which node of the two adjacent production nodes wants to tamper with the product image, it will be discovered, thereby ensuring that the product image of each production node is preserved intact and authentic. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood by reading the following detailed description with reference to the accompanying drawings. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein: Figure 1 1 is a flow chart schematically illustrating a method for tracing a seamless steel pipe continuous rolling production process according to the present invention; Figure 2 is a flowchart schematically illustrating step S3; Figure 3 It is a schematic diagram showing a pixel point as the center and a size equal to The neighborhood of The schematic diagram of the neighborhood in each direction is divided into Schematic diagram of the neighborhood in each direction; Figure 4 It is a schematic diagram showing a pixel point as the center and a size equal to The neighborhood of The schematic diagram of the neighborhood in each direction is divided into Schematic diagram of the neighborhood in each direction; Figure 5 It is a schematic diagram showing a pixel point as the center and a size equal to The neighborhood of The schematic diagram of the neighborhood in each direction is divided into Schematic diagram of the neighborhood in each direction. DETAILED DESCRIPTION

[0022] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work shall fall within the scope of protection of the present invention.

[0023] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0024] The seamless steel pipe continuous rolling production line is a highly automated and integrated production system. Each production node is closely connected to achieve high-efficiency and high-quality steel pipe production. The production nodes are as follows: 1. Raw material preparation node: Select appropriate steel as raw material, such as ingot, billet or continuous casting billet; 2. Raw material heating node: The raw materials are heated to an appropriate temperature for rolling. Heating is usually carried out in a heating furnace or induction furnace; 3. Raw material processing node: inspect the surface of the heated steel and remove the oxide scale and impurities on the surface; 4. Punching node: Use a punching machine to punch a hole in the center of the billet to form a tubular structure; 5. Rolling node: The pierced tube billet is rolled through a series of rolling mills to gradually reduce the outer diameter and wall thickness to form the required steel pipe size; 6. Sizing node: The outer diameter and wall thickness of the steel pipe are reduced by the sizing mill, and the outer diameter and wall thickness of the steel pipe are further accurately controlled by the sizing mill to achieve the final size requirements; 7. Cooling node: Cooling the rolled steel pipe to improve its internal structure and performance; 8. Straightening node: straighten the cooled steel pipe to ensure the straightness of the steel pipe; 9. Cutting nodes: Cut the steel pipe into the required length; 10. Surface treatment node: The surface of the steel pipe is pickled, shot blasted or coated to improve corrosion resistance and appearance quality.

[0025] Between different production nodes, due to production deviations, equipment failures, operational errors and other factors, the products produced may have defects in subsequent inspection and use.

[0026] When a product has quality problems, the responsibility is determined by tracing back the product images during the production process. However, if the product images are tampered with, the relevant production nodes can evade accountability. In order to improve the accuracy of accountability, it is necessary to ensure that the product images of each production node are preserved intact and authentic.

[0027] The embodiment of the present invention discloses a method for tracing the production process of seamless steel pipe continuous rolling, referring to Figure 1 , including steps S1 to S5: S1. Collect product images of each production node in the continuous rolling production process of seamless steel pipes, wherein the product images include multiple focus areas and one marked area.

[0028] Specifically, the seamless steel pipe continuous rolling production line is a highly automated and integrated production system. Each production node is closely connected. The production nodes include raw material preparation node, raw material heating node, raw material processing node, piercing node, rolling node, sizing node, cooling node, straightening node, cutting node, and surface treatment node, a total of 10 production nodes. The serial numbers of these 10 production nodes are 1 to 10, respectively. Production nodes to production nodes.

[0029] Furthermore, for the multiple production nodes included in the seamless steel pipe continuous rolling production process, any production node is taken as the target node, and the product image of the semi-finished product in the target node is collected; wherein, the semi-finished product in the target node refers to the semi-finished product processed by the previous production node of the target node; the product image of the semi-finished product in the target node includes the product image of the semi-finished product in the target node before being processed by the target node and the product image of the semi-finished product in the target node after being processed by the target node.

[0030] Furthermore, the product image includes multiple focus areas, and the areas in the product image other than all focus areas are recorded as annotated areas; wherein the focus areas included in the product image before processing are specified by the target node, and the focus areas included in the product image after processing are specified by the production node after the target node.

[0031] Among them, the focus area needs to be set according to the actual application scenario and needs. Therefore, in this embodiment, the focus area mainly refers to the main area in the product image of the semi-finished product processed by the previous production node that affects the product quality of the semi-finished product of the next production node. This needs to be specified by the staff of the next production node based on experience.

[0032] It should be noted that the focus area of ​​the product image in the present invention is the main area in the product image of the semi-finished product after processing by the previous production node that affects the product quality of the semi-finished product of the next production node. Therefore, the focus area in the product image before processing in the target node is specified by the target node, which can ensure that the main area affecting the product quality of the semi-finished product of the target node is focused on, thereby ensuring that when the target node is traced, the product image of the target node is preserved intact and authentic, and can more accurately determine whether it is the responsibility of the target node, thereby improving the accuracy of determining the responsibility based on tracing the product images during the production process.

[0033] S2. Set a key for each production node.

[0034] It should be noted that chaotic mapping functions are pseudo-random, sensitive to initial conditions, non-periodic and long-term unpredictable, and are often used as key generators.

[0035] Specifically, when the parameters and initial values ​​in the initial conditions of the two-dimensional chaotic mapping function are within their respective value ranges, the two-dimensional chaotic mapping function is in a chaotic state. Therefore, according to the initial conditions of the two-dimensional chaotic mapping function, a key is set for each production node, specifically: the parameters and initial values ​​in the initial conditions of the two-dimensional chaotic mapping function are evenly divided into two groups; for any production node, if the serial number of the production node is odd, the corresponding parameters and initial values ​​are randomly generated from the value range of the parameters and initial values ​​of the first group as the key of the production node; if the serial number of the production node is even, the corresponding parameters and initial values ​​are randomly generated from the value range of the parameters and initial values ​​of the second group as the key of the production node.

[0036] In one embodiment, a key is set for each production node according to the initial conditions of the two-dimensional Logistic chaotic mapping function, specifically: the initial conditions of the two-dimensional Logistic chaotic mapping function include 4 parameters and 2 initial values, and the parameters The value ranges are and ,parameter The value ranges are and , initial value and The value range of ; The parameters in the initial conditions of the two-dimensional Logistic chaotic mapping function ,parameter and initial values The first group is formed by taking the parameters in the initial conditions of the two-dimensional Logistic chaotic mapping function as ,parameter and initial values Form the second group; for any production node, if the serial number of the production node is odd, the parameters in the first group ,parameter and initial values Randomly generate a parameter within the range of ,parameter and initial values , as the key of the production node; if the serial number of the production node is an even number, the parameters in the second group ,parameter and initial values Randomly generate a parameter within the range of ,parameter and initial values , as the key of the production node.

[0037] In another embodiment, a key is set for each production node according to the initial conditions of the improved two-dimensional Henon chaotic mapping function, specifically: the initial conditions of the improved two-dimensional Henon chaotic mapping function include 2 parameters and 2 initial values, and when the parameters ,parameter , initial value and All in When the improved two-dimensional Henon chaotic mapping system is in a chaotic state, the two parameters of the improved two-dimensional Henon chaotic mapping function are fixed values, and the initial value is and The value range of ; The initial value of the initial condition of the improved two-dimensional Henon chaotic mapping function The first group is formed by converting the initial value of the initial condition of the improved two-dimensional Henon chaotic mapping function into Form the second group; for any production node, if the serial number of the production node is odd, the initial value in the first group Randomly generate an initial value within the range of , as the key of the production node; if the serial number of the production node is an even number, the initial value in the second group Randomly generate an initial value within the range of , as the key of the production node.

[0038] It should be noted that the present invention evenly divides the parameters and initial values ​​in the initial conditions of the two-dimensional chaotic mapping function into two groups, and distributes them to two adjacent production nodes respectively, wherein the first group is used as the key of the production node with an odd serial number, and the second group is used as the key of the production node with an even serial number; thereby ensuring that the keys of any two adjacent production nodes can be combined to form the complete initial conditions of the two-dimensional chaotic mapping function.

[0039] S3. Obtain first secret information from a focus area of ​​the product image according to the keys of each production node and its adjacent production nodes.

[0040] Take any production node as the target node, and obtain the first secret information from the focus area of ​​the product image according to the key of the target node and its adjacent production nodes; therefore, the flowchart of step S3 refers to Figure 2 , including steps S301 to S304, specifically: S301 : For a pixel point in a target area, obtain a feature value of the pixel point in each direction.

[0041] Specifically, for any pixel in the focus area of ​​the product image, a pixel with a size equal to The neighborhood of the pixel is evenly divided to obtain the pixel point in Direction to The pixel point itself does not belong to the neighborhood of the pixel point in any direction, so the first Direction to The number of all pixels in the neighborhood of the direction is equal to , are respectively a preset first value and a preset second value.

[0042] Among them, the preset first value The specific value of can be set according to the actual application scenario and requirements, and for The present invention will be within the singular range Set to 5; the default second value The specific value of can be set according to the actual application scenario and requirements, and is 4 or 8, the present invention will Set to 8.

[0043] For example, Figure 3 The pixel is centered and equal to Neighborhood, Figure 3 (1) in the formula is The neighborhood is divided into Neighborhood in directions, where All the pixels make up the Neighborhood of direction, for Integers in the range, for example, all pixels marked as 1 constitute the Neighborhood of direction; Figure 3 (2) in the formula is The neighborhood is divided into Neighborhood in directions, where All the pixels make up the Neighborhood of direction, for Integers in the range, for example, all pixels marked with 5 constitute the neighborhood in the 5th direction.

[0044] For example, Figure 4 The pixel is centered and equal to Neighborhood, Figure 4 (1) in the formula is The neighborhood is divided into Neighborhood in directions, where All the pixels make up the Neighborhood of direction, for Integers in the range, for example, all pixels marked as 2 constitute the Neighborhood of direction; Figure 4 (2) in the formula is The neighborhood is divided into Neighborhood in directions, where All the pixels make up the Neighborhood of direction, for Integers in the range, for example, all pixels marked as 6 constitute the neighborhood in the 6th direction.

[0045] For example, Figure 5 The pixel is centered and equal to Neighborhood, Figure 5 (1) in the formula is The neighborhood is divided into Neighborhood in directions, where All the pixels make up the Neighborhood of direction, for Integers in the range, for example, all pixels marked as 4 constitute the Neighborhood of direction; Figure 5 (2) in the formula is The neighborhood is divided into Neighborhood in directions, where All the pixels make up the Neighborhood of direction, for Integers in the range, for example, all pixels marked as 8 constitute the neighborhood in the 8th direction.

[0046] Furthermore, based on the difference between the grayscale value of the pixel and the grayscale values ​​of other pixels in the neighborhood of the pixel in each direction, the characteristic value of the pixel in each direction is calculated to satisfy the expression: ; Where, For this pixel point The eigenvalue of the direction, is the grayscale value of the pixel, For this pixel point The mean of the grayscale values ​​of all other pixels in the neighborhood of the direction, To take the absolute value, for Integers in the range, is the preset second value.

[0047] It should be noted that Indicates the grayscale value of the pixel With the pixel point The mean of the grayscale values ​​of all other pixels in the neighborhood of the direction The present invention obtains the characteristic values ​​of the pixels in various directions based on the local characteristics of the pixels in the area of ​​interest. When the grayscale values ​​of the pixels in the area of ​​interest are tampered with, the local characteristics of the related pixels will change, which will cause the characteristic values ​​of the related pixels in various directions to change as well. This ensures that the product images of each production node are preserved intact and authentic. When quality problems occur in subsequent products, the accuracy of determining responsibility through tracing based on product images during the production process is improved.

[0048] S302 : Obtain segmentation thresholds in each direction according to the distribution of the characteristic values ​​of all historical pixels in the historically annotated area of ​​the historical image of the target node in each direction.

[0049] Specifically, all historical images of the target node are obtained; for the Direction to In any direction, the characteristic values ​​of all historical pixel points in the focus area of ​​the historical image in that direction are obtained; according to the size relationship between the characteristic values ​​of the historical pixel points in that direction and the separation value, all historical pixel points are divided into two categories, including: all historical pixel points with characteristic values ​​in that direction less than the separation value are divided into the first category, and all historical pixel points with characteristic values ​​in that direction greater than or equal to the separation value are divided into the second category.

[0050] Furthermore, the range of the separation value is , Calculate different separation values ​​for the minimum and maximum values ​​of the characteristic values ​​of all historical pixel points in this direction respectively. The preference , satisfying the expression: ; Where, is the intra-class variance of the eigenvalues ​​of all historical pixels in the first category in this direction, is the intra-class variance of the eigenvalues ​​of all historical pixels in the second category in this direction, is the inter-class variance of the eigenvalues ​​of all historical pixels in the first and second categories in this direction, is the natural exponential function, is the mean of the characteristic values ​​of all historical pixels in the category with the least number of historical pixels in this direction, To take the absolute value.

[0051] The calculation methods of the intra-class variance and the inter-class variance are well-known technologies and will not be described in detail here.

[0052] Furthermore, the separation value with the highest preference is used as the segmentation threshold in this direction.

[0053] It should be noted that, according to the relationship between the characteristic value of the historical pixel point in this direction and the separation value, all historical pixel points are divided into two categories. Among the two categories obtained, one category represents the distribution of the characteristics of the product images of products with good quality, and the other category represents the distribution of the characteristics of the product images of products with poor quality. Since the proportion of products with poor quality is relatively low, the number of categories representing products with poor quality is the least. When calculating the preference of different separation values, when the separation value is The smaller the difference between the mean of the characteristic values ​​of all historical pixels in the category with the least number of historical pixels in this direction, the smaller the separation value. The preference The larger the value, the closer the final selected segmentation threshold is to the characteristic value of the poor quality product. This ensures that when you want to tamper with the product image of the poor quality product in the future, the characteristic identification of the pixel point is more easily destroyed, making the tampering easier to detect.

[0054] S303 : For the pixel points in the focus area, obtain the feature identifier of the pixel points according to the relationship between the feature values ​​of the pixel points in each direction and the segmentation thresholds in each direction.

[0055] Specifically, for any pixel in the region of interest, a feature identifier of the pixel is obtained based on the relationship between the feature value of the pixel in each direction and the segmentation threshold in each direction; the feature identifier is a binary number and satisfies the expression: ; Where, is the first feature identifier of the pixel point bits, For this pixel point The eigenvalue of the direction, For the The segmentation threshold of the direction, for Integers in the range, is the preset second value.

[0056] It should be noted that the present invention obtains the feature identification of the pixel points based on the local features of the pixel points in the area of ​​interest. When the grayscale value of the pixel points in the area of ​​interest is tampered with, the local features of the relevant pixel points will change, which in turn causes the feature identification of the relevant pixel points to change. This ensures that the product images of each production node are preserved intact and authentic. When quality problems occur in subsequent products, the accuracy of determining responsibility based on tracing the product images during the production process is improved.

[0057] S304: Perform bitwise operations on the feature identifiers of all pixels according to the secret keys of the target node and its adjacent production nodes to obtain first secret information.

[0058] Specifically, for the target node, the production nodes adjacent to it include the previous production node of the target node and the next production node of the target node. Therefore, for the product image before processing in the target node, the key of the target node and the production node adjacent to it refers to performing bit operations on the feature identifiers of all pixels in the focus area of ​​the product image based on the key of the target node and the key of the previous production node of the target node to obtain the first secret information; for the product image after processing in the target node, the key of the target node and the production node adjacent to it refers to performing bit operations on the feature identifiers of all pixels in the focus area of ​​the product image based on the key of the target node and the key of the next production node of the target node to obtain the first secret information.

[0059] Furthermore, based on the secret keys of the target node and its adjacent production nodes, bitwise operations are performed on the characteristic identifiers of all pixels to obtain the first secret information, including: 1. Take the key of the target node and its adjacent production nodes as the initial conditions of the two-dimensional chaotic mapping function, and iterate the two-dimensional chaotic mapping function. times, and each iteration obtains a chaotic value pair; each chaotic value pair consists of two chaotic values, and the two chaotic values ​​are the first chaotic value and the second chaotic value, respectively. Equal to the number of all pixels in the region of interest.

[0060] 2. Obtaining the first secret information requires Secondary operation; among them, , proceed to When performing the second-order operation, The result of the secondary operation Bit and The feature identifier of the pixel The bits are aligned and then bitwise operations are performed.

[0061] The bit operations include but are not limited to XOR operations and XNOR operations, which are well-known technologies and will not be described in detail here. In this embodiment, the bit operation refers to the XNOR operation.

[0062] and Satisfies the expression: ; ; Where, Respectively The first chaotic value and the second chaotic value in a chaotic value pair, For the The number of all bits in the result of the bitwise operation and the number of all bits in the feature identifiers of all pixels are equal to , is the preset second value.

[0063] Special note is that when , that is, to carry out the When performing the second-order operation, The feature identifier of the pixel Bit and The feature identifier of the pixel The bits are aligned and then bit operations are performed, where , Where, Respectively The first chaotic value and the second chaotic value in the chaotic value pair, the number of all bits in the feature identifier of all pixels is equal to , is the preset second value.

[0064] 3. The result of the secondary operation is used as the first secret information.

[0065] It should be noted that the present invention uses the keys of two adjacent production nodes associated with the product image to perform bit operations on the feature identifiers of all pixels in the area of ​​interest to obtain the first secret information. This can ensure that no matter which of the two adjacent production nodes wants to tamper with the product image, it will be discovered, thereby ensuring that the product image of each production node is preserved intact and authentic.

[0066] S4. Embed the first secret information into the marked area using a steganographic algorithm to obtain a secret image.

[0067] Specifically, the first secret information is embedded into the marked area of ​​the product image through the EMD steganography algorithm to obtain a secret image; the EMD steganography algorithm is a well-known technology and will not be described in detail here.

[0068] Among them, the EMD steganography algorithm must first be established The size of the magic cube matrix is ​​because the grayscale value of the pixel in the grayscale image ranges from , so the value ranges of the X-axis and Y-axis of the magic cube matrix are , and only contains numbers 0 to 4, and if you randomly select a cross-shaped area (occupying five grids) in this matrix, the numbers you get are 0, 1, 2, 3, and 4.

[0069] Furthermore, for a pair of consecutive pixels in the marked area, the grayscale values ​​of the two pixels in the pair are used as the X-axis and Y-axis to obtain a point in the matrix, and a cross is taken with this point as the center point; for the pair of consecutive pixels, the hidden information is two bits in the first secret information; the hidden information is converted into a hidden information number in base 40, and the cross is moved in the direction of the cross (i.e., up, down, left, right, or middle) where the hidden information number is, so as to obtain the changed grayscale value of the pair of consecutive pixels; for a pair of consecutive pixels (i.e., two pixels) taken by this method, only the grayscale value of one of the pixels will change, which is the grayscale value of the pixel plus or minus one, and the grayscale value of the other pixel remains unchanged. It is also possible that the grayscale values ​​of both pixels remain unchanged.

[0070] It should be noted that due to the particularity of the magic cube matrix in the EMD steganography algorithm, the cross matrix in the EMD steganography algorithm must contain five values ​​​​from 0 to 4, so that adding or subtracting one from the grayscale value of the pixel point can achieve the hiding of two bits; therefore, the EMD steganography algorithm makes very little modification to the grayscale value of the pixel value in the marked area, making it more difficult to be discovered, and at the same time will not affect the image quality of the marked area.

[0071] S5. When tracing the production process, in response to the verification sequence being identical to the second secret information, the product image of the traced node is preserved intact and authentic.

[0072] Specifically, when tracing the production process, bit operations are performed on the feature identifiers of all pixels in the focus area of ​​the secret image based on the keys of the traced node and its adjacent production nodes to obtain the second secret information; a verification sequence is extracted from the marked area of ​​the secret image through a steganographic algorithm; in response to the verification sequence being identical to the second secret information, the product image of the traced node is preserved complete and authentic.

[0073] Among them, for the traced node, if the secret image is the secret image corresponding to the product image before processing in the traced node, the method of performing a bit operation on the feature identifiers of all pixels in the focus area of ​​the secret image based on the key of the traced node and the key of the production node immediately preceding the traced node to obtain the second secret information according to the key of the traced node and the key of the production node immediately preceding the traced node. If the secret image is the secret image corresponding to the product image after processing in the traced node, the method of performing a bit operation on the feature identifiers of all pixels in the focus area of ​​the secret image based on the key of the traced node and the key of the production node immediately preceding the traced node to obtain the second secret information according to the key of the traced node and the key of the production node immediately preceding the traced node.

[0074] Among them, the extraction of the verification sequence from the marked area of ​​the secret image through the EMD steganography algorithm is the inverse process of embedding the first secret information into the marked area of ​​the product image through the EMD steganography algorithm; and the EMD steganography algorithm is a well-known technology and will not be described here in detail.

[0075] It should be noted that the focus area of ​​the product image is the main area in the product image of the semi-finished product processed by the previous production node that affects the product quality of the semi-finished product of the next production node. Therefore, the first secret information is obtained from the focus area of ​​the product image and embedded in the marked area of ​​the product image. When tracing the production process, when the focus area that affects the product quality is tampered with, the second secret information obtained from the focus area of ​​the secret image is different from the verification sequence extracted from the marked area of ​​the secret image, thereby ensuring that the product images of each production node are preserved intact and authentic. When quality problems occur in subsequent products, the accuracy of determining the responsibility based on the product images in the production process is improved. Among them, since the product image is associated with two adjacent production nodes, it is an important basis for dividing the responsibility between the two adjacent production nodes. The present invention obtains the first secret information from the focus area of ​​the product image according to the key of the target node and the production node adjacent to it, which can ensure that no matter which node of the two adjacent production nodes wants to tamper with the product image, it will be discovered, thereby ensuring that the product image of each production node is preserved intact and authentic.

[0076] In the description of this specification, "multiple" and "several" mean at least two, such as two, three or more, unless otherwise clearly defined.

[0077] While this specification has shown and described several embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Numerous modifications, variations, and alternatives will occur to those skilled in the art without departing from the concept and spirit of the present invention. It should be understood that in practicing the present invention, alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. A method for tracing the production process of seamless steel pipe continuous rolling, characterized in that: include: For a target node among multiple production nodes in the seamless steel pipe continuous rolling production process, a product image of a semi-finished product in the target node is collected, and the product image includes multiple focus areas and one marked area; According to the secret keys of the target node and the production nodes adjacent thereto, the first secret information is obtained from the focus area of ​​the product image, including: obtaining the segmentation threshold in each direction according to the distribution of the characteristic values ​​of all historical pixels in the focus area of ​​the historical image of the target node in each direction; for the pixels in the focus area, obtaining the characteristic identifier of the pixel according to the size relationship between the characteristic value of the pixel in each direction and the segmentation threshold in each direction, wherein the characteristic value of the pixel in each direction refers to the difference between the gray value of the pixel and the gray value of other pixels in the neighborhood of the pixel in each direction, and the characteristic identifier is a binary number; according to the secret keys of the target node and the production nodes adjacent thereto, bitwise operations are performed on the characteristic identifiers of all pixels to obtain the first secret information; Embedding the first secret information into the marked area by a steganographic algorithm to obtain a secret image; When tracing the production process, bit operations are performed on the feature identifiers of all pixels in the focus area of ​​the secret image according to the keys of the traced node and the production nodes adjacent to it to obtain the second secret information; a verification sequence is extracted from the marked area of ​​the secret image through a steganographic algorithm; in response to the verification sequence being the same as the second secret information, the product image of the traced node is preserved intact and authentic.

2. A seamless steel pipe continuous rolling production process tracing method according to claim 1, characterized in that: The multiple production nodes in the seamless steel pipe continuous rolling production process include raw material preparation nodes, raw material heating nodes, raw material processing nodes, perforation nodes, rolling nodes, sizing nodes, cooling nodes, straightening nodes, cutting nodes, and surface treatment nodes.

3. The method for tracing the production process of seamless steel pipe continuous rolling according to claim 1, characterized in that: The semi-finished product in the target node refers to the semi-finished product processed by the previous production node of the target node; the product image of the semi-finished product in the target node includes the product image of the semi-finished product in the target node before being processed by the target node and the product image of the semi-finished product in the target node after being processed by the target node; the area of ​​interest included in the product image before processing is specified by the target node, and the area of ​​interest included in the product image after processing is specified by the next production node of the target node.

4. The method for tracing the production process of seamless steel pipe continuous rolling according to claim 1, characterized in that: The key of the production node is set according to the initial condition of the two-dimensional chaotic mapping function, including: The parameters and initial values ​​in the initial conditions of the two-dimensional chaotic mapping function are evenly divided into two groups; For any production node, if the serial number of the production node is odd, the corresponding parameters and initial values ​​are randomly generated from the value range of the parameters and initial values ​​of the first group as the key of the production node; if the serial number of the production node is even, the corresponding parameters and initial values ​​are randomly generated from the value range of the parameters and initial values ​​of the second group as the key of the production node.

5. The method for tracing the production process of seamless steel pipe continuous rolling according to claim 1, characterized in that: The step of obtaining the segmentation threshold in each direction includes: Get all the historical images of the target node; for the Direction to In any direction, the characteristic values ​​of all historical pixels in the focus area of ​​the historical image in that direction are obtained; the range of the separation value is , are the minimum and maximum values ​​of the characteristic values ​​of all historical pixels in the direction, respectively; according to the size relationship between the characteristic values ​​of the historical pixels in the direction and the separation value, all historical pixels are divided into two categories, including: all historical pixels with characteristic values ​​in the direction less than the separation value are divided into the first category, and all historical pixels with characteristic values ​​in the direction greater than or equal to the separation value are divided into the second category; Calculate the preference of different separation values, and use the separation value with the largest preference as the segmentation threshold in that direction.

6. A seamless steel pipe continuous rolling production process tracing method according to claim 5, characterized in that: The preference of the different separation values ​​satisfies the expression: ； In the formula, For separation value The preference of are the intra-class variances of the feature values ​​of all historical pixels in the first and second categories in this direction, is the inter-class variance of the feature values ​​of all historical pixels in the first and second categories in this direction, is the natural exponential function, is the mean of the characteristic values ​​of all historical pixels in the category with the least number of historical pixels in this direction, To take the absolute value.

7. The method for tracing the production process of seamless steel pipe continuous rolling according to claim 1, characterized in that: The feature identifier of the pixel point The bits satisfy the expression: ； In the formula, is the feature identifier of the pixel point Bits, For this pixel point The eigenvalue of the direction, For the The segmentation threshold of the direction, for Integer in the range, is the preset second value.

8. The method for tracing the production process of seamless steel pipe continuous rolling according to claim 1, characterized in that: The characteristic values ​​of the pixel points in each direction include: For any pixel in the region of interest, The neighborhood of the pixel is evenly divided to obtain the pixel point in Direction to Neighborhood of direction, are respectively a preset first value and a preset second value; The pixel point is Eigenvalue of direction Satisfies the expression: ； In the formula, is the gray value of the pixel, For this pixel point The mean grayscale value of all other pixels in the neighborhood of the direction, To take the absolute value, for Integer in the range, is the preset second value.

9. The method for tracing the production process of seamless steel pipe continuous rolling according to claim 1, characterized in that: The obtaining of the first secret information comprises: The keys of the target node and its adjacent production nodes are used as the initial conditions of the two-dimensional chaotic mapping function, and the two-dimensional chaotic mapping function is iterated. times, and each iteration obtains a chaotic value pair. Respectively The first chaotic value and the second chaotic value in a chaotic value pair; Obtaining the first secret information requires Secondary operation, is equal to the number of all pixels in the focus area. The result of the secondary operation is used as the first secret information; Conduct the When performing the next bit operation, , will The result of the next operation is Bit and The feature identifier of the pixel The bits are aligned and then bit operations are performed; , , For the The number of all bits in the result of the bitwise operation and the number of all bits in the feature identifiers of all pixels are equal to , is the preset second value.

10. The method for tracing the production process of seamless steel pipe continuous rolling according to claim 3, characterized in that: The method further comprises: For the target node, the production nodes adjacent to it include the previous production node of the target node and the next production node of the target node. Therefore, for the product image in the target node before processing, the key based on the target node and the production nodes adjacent to it refers to the key based on the target node and the key of the previous production node of the target node; for the product image in the target node after processing, the key based on the target node and the production nodes adjacent to it refers to the key based on the target node and the key of the next production node of the target node.

Images