Strip steel positioning method and device

Abstract

The invention relates to the field of industrial application of image recognition, in particular to a strip steel position positioning method and device, and the method comprises the steps: obtaining a strip steel image of a segment region; extracting a band end mask and a band side mask in the band steel image based on semantic segmentation; respectively calculating a first horizontal coordinate mean value and a second horizontal coordinate mean value of the belt end part mask and the belt side surface mask under the condition that the belt end part mask and the belt side surface mask meet a preset condition; carrying out subtraction on the first horizontal coordinate mean value and the second horizontal coordinate mean value to obtain a first difference value; threshold judgment is conducted on the first difference value to determine whether the section area is the strip steel head or the strip steel tail; and obtaining corresponding position information of the strip steel head and the strip steel tail through the strip end mask. The automation and precise positioning of the whole process can help to improve the production efficiency and quality of the strip steel, and the accurate head and tail position information helps to avoid misoperation and reduce the generation of waste products and inferior-quality products, so that the efficiency and economic benefits of a production line are improved.

Description

Technical Field

[0001] The present application relates to the field of industrial application of image recognition, and in particular to a method and device for locating the position of a steel strip. Background Art

[0002] The purpose of strip steel transportation in the steel plant production process is to transport steel from one place to another for subsequent processing and production.

[0003] The steel head refers to the front part of the strip, which is also the starting position of the strip coil; during the transmission of the strip, the position of the strip head needs to be monitored and controlled to ensure the smooth progress of subsequent processing steps such as shearing and rolling. The accurate position information of the strip head is very important to operators and equipment, because they need to locate, control and start the corresponding production process according to the area and position of the strip head. Therefore, the position monitoring of the strip head is a very critical part of the steel plant production process. If the position of the strip head cannot be monitored in real time during the transmission of the strip in the steel plant production process, it may cause the subsequent strip shearing and rolling production to stagnate. At the same time, the failure to correctly and timely monitor the position of the strip head is also prone to safety accidents. The strip shearing equipment and operators need to obtain the current area and position of the strip head to quickly locate the problems of the strip jamming and starting, and to start the processing steps of the strip at different positions of the transmission belt in sequence. If workers are allowed to observe the position of the strip head at close range, there will be safety risks.

[0004] Therefore, how to provide a general solution that can ensure real-time tracking of the strip head position while having both processing speed and complex feature processing capabilities is the technical problem to be solved by this application. Summary of the invention

[0005] In order to overcome the above technical problems, the present application provides a method and device for locating the position of a steel strip.

[0006] According to a first aspect of the present application, a method for locating a strip position is provided, comprising: acquiring a strip image of a segment area; extracting a strip end mask and a strip side mask in the strip image based on semantic segmentation; respectively calculating a first horizontal coordinate mean and a second horizontal coordinate mean of the strip end mask and the strip side mask when the strip end mask and the strip side mask meet preset conditions; subtracting the first horizontal coordinate mean and the second horizontal coordinate mean to obtain a first difference; performing a threshold judgment on the first difference to determine whether the segment area is a strip head or a strip tail; and acquiring position information corresponding to the strip head and the strip tail through the strip end mask.

[0007] In a further scheme of the present application, the strip end mask and the strip side mask in the strip image are extracted based on semantic segmentation, including: defining category labels for the strip ends, strip sides, and background, respectively; performing semantic segmentation on the strip image according to the category labels through a pre-trained semantic segmentation model, and extracting a mask image in the prediction result; and extracting a strip end mask and a strip side mask from the mask image.

[0008] In a further scheme of the present application, the first horizontal coordinate mean and the second horizontal coordinate mean according to the end mask and the side mask include: traversing each pixel with the end mask and the side mask respectively, recording all the first position coordinates and the second position coordinates, adding the horizontal coordinates of the first position coordinates and dividing them by the number of pixels of the first position coordinates to obtain the first horizontal coordinate mean; adding the horizontal coordinates of the second position coordinates and dividing them by the number of pixels of the second position coordinates to obtain the second horizontal position coordinate mean.

[0009] In a further scheme of the present application, after extracting the strip end mask and the strip side mask in the strip image based on semantic segmentation, it also includes: respectively obtaining the strip end bounding box and the strip side bounding box of the strip end mask and the strip side mask; obtaining the intersection-union ratio of the strip end bounding box and the strip side bounding box; judging whether the intersection-union ratio is within a preset threshold range, wherein the threshold range is from a first threshold to a second threshold; when the intersection-union ratio is less than the first threshold, outputting that the segment area has no strip end; when the intersection-union ratio is greater than the second threshold, outputting that the segment area is in a transition stage; and when the intersection-union ratio is between the first threshold and the second threshold, determining that the strip end mask and the strip side mask meet preset conditions.

[0010] In a further scheme of the present application, an end bounding box and a side bounding box with an end mask and a side mask are obtained respectively, including: performing morphological corrosion on the end mask and the side mask respectively to obtain a steel head fine mask and a side fine mask; and obtaining an end bounding box and a side bounding box by calculating a connected domain bounding box for the steel head fine mask and the side fine mask.

[0011] In a further scheme of the present application, after extracting the strip end mask and the strip side mask in the strip steel image based on semantic segmentation, it also includes: counting the number of pixels of the strip end mask; calculating the strip end area according to the number of pixels of the strip end mask; judging whether the strip end area is within a preset area range, wherein the area range includes a first area threshold to a second area threshold; when the strip end area is less than the preset first area threshold, outputting that the segment area has no strip end; when the strip end area is greater than the preset second area threshold, outputting that this segment area is in a transition stage; and when the strip end area is between the first area threshold and the second area threshold, determining that the strip end mask and the strip side mask meet preset conditions.

[0012] In a further solution of the present application, the positioning method further includes: drawing the position of the strip head according to the strip end bounding box, wherein the strip head position includes an outer rectangle and a center point; and outputting the strip head position.

[0013] In a further solution of the present application, judging whether the segment area is the strip head or the strip tail according to the first difference includes: when the first difference is greater than 0, determining that the segment area is the strip head; when the first difference is less than 0, determining that the segment area is the strip tail.

[0014] In a further scheme of the present application, the positioning method also includes: accumulating and counting once to obtain a first value each time the position information of the strip head is obtained; calculating once to obtain a second value each time the position information of the strip tail is obtained; when the second difference between the first value and the second value is greater than 1 or less than 0, the output section area is abnormal.

[0015] In summary, the present application provides a method based on semantic segmentation and coordinate calculation to obtain the position information of the strip head and the strip tail. By using semantic segmentation technology, the strip end mask and the strip side mask in the strip image can be automatically extracted, avoiding the tediousness and subjectivity of manual processing. Then, by calculating the mean of the horizontal coordinates of the strip end mask and the strip side mask, the position information of the strip head and the strip tail can be obtained, and it is determined whether the segment area is the strip head or the strip tail. After obtaining the position information of the strip head and the strip tail, the position information of the strip head and the strip tail can be further obtained through the mask to facilitate subsequent data analysis and processing according to the position information. For example, the position information of the strip head and the strip tail can be statistically obtained to obtain the length of the strip, or whether the strip is kept straight during operation, etc., to provide data support for the optimization and quality control of the production process. The automation and precise positioning of the entire process can help improve the efficiency and quality of strip production. Accurate head and tail position information helps to avoid wrong operations, reduce the generation of waste and defective products, thereby improving the efficiency and economic benefits of the production line.

[0016] Other features and advantages of the embodiments of the present invention will be described in the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the specific implementation methods of the present application or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 A flow chart of a method for locating a steel strip provided by an embodiment of the present invention;

[0019] Figure 2 A logical topological diagram of an overall specific solution of the strip steel position positioning method provided by an embodiment of the present invention;

[0020] Figure 3 A schematic flow chart of step S20 in the method for locating the position of a steel strip provided in an embodiment of the present invention;

[0021] Figure 4 A real-life picture of the position of the strip provided by an embodiment of the present invention;

[0022] Figure 5 A schematic flow chart of step S30 in the method for locating the position of a steel strip provided in an embodiment of the present invention;

[0023] Figure 6 A schematic flow chart of step S70 in the method for locating the position of a steel strip provided in an embodiment of the present invention;

[0024] Figure 7 A schematic flow chart of step S71 in the method for locating the position of a steel strip provided in an embodiment of the present invention;

[0025] Figure 8 A schematic flow chart of step S80 in the method for locating the position of a steel strip provided in an embodiment of the present invention;

[0026] Fig. 9 Schematic diagrams that simply represent the strip images under various working conditions;

[0027] Fig.10 A schematic flow chart of step S50 in the method for locating the position of a steel strip provided in an embodiment of the present invention;

[0028] Fig.11 A schematic flow chart of step S60 in the method for locating the position of a steel strip provided in an embodiment of the present invention;

[0029] Fig.12 A schematic flow chart of step S90 in the method for locating the position of a steel strip provided in an embodiment of the present invention; and

[0030] Fig.13 A module schematic diagram of a positioning system with a steel head position is provided.

[0031] Marking Description

[0032] 100. Positioning system;

[0033] 10. Image acquisition module; 20. Extraction module;

[0034] 30. A first calculation module; 40. A second calculation module;

[0035] 50. Judgment module; 60. Output module. DETAILED DESCRIPTION

[0036] In order to make the above and other features and advantages of the present application more clear, the present application is further described below in conjunction with the accompanying drawings. It should be understood that the specific embodiments given herein are for the purpose of explaining to those skilled in the art and are only exemplary and not restrictive.

[0037] In the following description, many specific details are set forth to provide a thorough understanding of the present application. However, it is apparent to those skilled in the art that specific details need not be adopted to practice the present application. In other cases, well-known steps or operations are not described in detail to avoid blurring the present application.

[0038] The method for locating the position of the steel strip provided in the embodiment of the present application can be executed by the data detection device provided in the embodiment of the present application, and the device can be configured in an electronic device.

[0039] See also Figure 1 , Figure 1 The present invention provides a flow chart of a strip steel location positioning method. The strip steel location positioning method includes:

[0040] Step S10, obtaining a strip steel image of a segment area;

[0041] Step S20, extracting a strip end mask and a strip side mask in the strip image based on semantic segmentation;

[0042] Step S30: when the band end mask and the band side mask meet the preset conditions, respectively calculate the first abscissa mean and the second abscissa mean of the band end mask and the band side mask;

[0043] Step S40, subtracting the first horizontal coordinate mean from the second horizontal coordinate mean to obtain a first difference;

[0044] Step S50, performing a threshold judgment on the first difference to determine whether the segment area is a strip head or a strip tail;

[0045] Step S60: obtaining position information corresponding to the strip head and the strip tail through the strip end mask.

[0046] Specifically, step S10 uses a side alignment camera installed on one side of the transmission belt to collect a motion image of the strip head. In the embodiment of the present invention, the motion image of the strip head is referred to as a strip image.

[0047] In step S20, a semantic segmentation algorithm is used to process the above-mentioned strip image, extract the strip end and the strip side, and generate corresponding masks. The strip end mask and the strip side mask are both binary images. For example, the strip end area is marked as 1, and the strip side area is marked as 0.

[0048] Those skilled in the art can understand that the strip is a rectangle after it is just unfolded. In the present invention, the long side of the strip is interpreted as the "strip side" mentioned above, and the short side is interpreted as the "strip end" mentioned above.

[0049] It is understandable that due to the small area of ​​the strip end, false detection is prone to occur at the model level. Therefore, in post-processing, the strip side is used to filter possible false detections of the strip head to improve its accuracy and distinguish the strip head and strip tail.

[0050] In addition, since the strip head moves very quickly, if the traditional machine vision solution is used based on the characteristics of the strip head color, strip shape, etc., it is often necessary to manually set the algorithm threshold parameters, and the manual adjustment and optimization workload is large. In addition, it is sensitive to factors such as the light, shadow, and color range of the transmission belt environment and is easily disturbed, resulting in low yield and versatility. Similar problems also exist in low-information sensing mechanisms such as laser sensors and pressure sensors. Therefore, this application uses semantic segmentation to extract the strip end mask and strip side mask in the strip image. Semantic segmentation can classify each pixel in the image, not just relying on local color and shape features. This can enhance the algorithm's adaptability to various lighting conditions and shadows, and improve the algorithm's robustness and versatility.

[0051] Step S30 is to calculate the average value of the horizontal coordinates (x coordinates) of all pixels in the strip end mask and the strip side mask. By calculating the average value, the approximate position of the strip head and the strip side can be obtained.

[0052] In step S40 to step S50, a simple judgment can be made to determine whether the segment area is the head of the strip or the tail of the strip according to the size of the first difference. The specific judgment conditions can be determined according to the actual situation. For example, when the first difference is greater than a certain threshold, it can be determined as the head of the strip, otherwise it is determined as the tail of the strip. After determining whether the segment area is the head of the strip or the tail of the strip, step S60 then obtains the position information of the head of the strip and the tail of the strip through the strip end mask, and outputs the position information of the head of the strip and the tail of the strip together with the steel coil information, such as: steel coil number-strip head position-strip tail position. This provides accurate data for subsequent automated processing technology, establishes a data model for strip production, and further optimizes the production process and improves production efficiency.

[0053] In summary, the present invention provides a method based on semantic segmentation and coordinate calculation to obtain the position information of the strip head and the strip tail. By using the semantic segmentation technology, the strip end mask and the strip side mask in the strip image can be automatically extracted, avoiding the tediousness and subjectivity of manual processing. Then, by calculating the mean of the horizontal coordinates of the strip end mask and the strip side mask, the position information of the strip head and the strip tail can be obtained, and it is determined whether the segment area is the strip head or the strip tail. After obtaining the position information of the strip head and the strip tail, the position information of the strip head and the strip tail can be further obtained through the mask to facilitate subsequent data analysis and processing according to the position information. For example, the position information of the strip head and the strip tail can be statistically obtained to obtain the length of the strip, or whether the strip is kept straight during operation, etc., to provide data support for the optimization and quality control of the production process. The automation and precise positioning of the entire process can help improve the efficiency and quality of strip production. Accurate head and tail position information helps to avoid wrong operations and reduce the generation of waste and defective products, thereby improving the efficiency and economic benefits of the production line.

[0054] In a specific solution of an embodiment of the present invention, a semantic segmentation technology based on deep learning is used to solve the problem of real-time tracking of the belt end positions in various areas of the transmission belt, and the EfficientNetV3 model is used as the neural network backbone to obtain the belt end mask and the belt side mask.

[0055] Please combine Figure 3 , Figure 3 The flowchart of step S20 in the method for locating the position of a steel strip provided in an embodiment of the present invention is as follows. In an embodiment of the present invention, extracting a strip end mask and a strip side mask in a steel strip image based on semantic segmentation includes:

[0056] Step S21, respectively define the category labels of the band end, band side, and background;

[0057] Step S22, performing semantic segmentation on the steel strip image according to the category label using a pre-trained semantic segmentation model, and extracting a mask image in the prediction result;

[0058] Step S23: Create strip head and strip side masks from the mask image.

[0059] It can be understood that the category labels defining the ends, sides, and background respectively can be labeled with integer values, such as the end category is represented by an integer value of 1, the side category is represented by an integer value of 2, and the background category is represented by an integer value of 0; or color-coded labels can be used, such as the end category is red (RGB value [255,0,0]), the side category is green (RGB value [0,255,0]), and the background is (RGB value [0,0,255]).

[0060] Use a pre-trained semantic segmentation model, such as the EfficientNetV3 model, to input the strip image into the semantic segmentation model and obtain the prediction result. The prediction result is a mask image with the same size as the original image, in which each pixel is predicted to be one of the strip ends, strip sides, or background. Then, the mask image in the prediction result is divided into three parts: strip ends, strip sides, and background according to the category label. For example, the pixel value of the strip ends in the prediction result is set to 1, and the pixel values ​​of other areas are set to 0. In this way, the strip end mask and the strip side mask are segmented, such as Figure 4 Actual image of the demonstration shown.

[0061] Please continue reading Figure 5 , Figure 5 The schematic diagram of the flow chart of step S30 in the strip steel position positioning method provided in the embodiment of the present invention. The first horizontal coordinate mean and the second horizontal coordinate mean according to the strip end mask and the strip side mask include:

[0062] Step S31, traverse each pixel with an end mask and each pixel with a side mask respectively, and record all first position coordinates and second position coordinates;

[0063] Step S32, adding the horizontal coordinates of the first position coordinates and dividing the sum by the number of pixels of the first position coordinates to obtain a first horizontal coordinate mean;

[0064] Step S33: add the horizontal coordinates of the second position coordinates and divide the sum by the number of pixels of the second position coordinates to obtain a second horizontal position coordinate mean.

[0065] These steps are mainly for processing and calculating pixels with end masks and side masks. First, traverse all pixels and record the first position coordinates and the second position coordinates respectively. Then, add the horizontal coordinates of the first position and divide them by the number of pixels to obtain the first horizontal coordinate mean. Similarly, add the horizontal coordinates of the second position and divide them by the number of pixels to obtain the second horizontal coordinate mean.

[0066] The first abscissa mean and the second abscissa mean may characterize the center positions of the belt ends and the belt sides.

[0067] Please combine 6, Figure 6 This is a schematic flow chart of step S70 in the method for locating the position of a steel strip provided in an embodiment of the present invention.

[0068] Positioning methods also include:

[0069] Step S71, respectively obtaining an end bounding box with an end mask and a side bounding box with a side mask;

[0070] Step S72, obtaining the intersection-union ratio of the end bounding box and the side bounding box;

[0071] Step S73, determining whether the intersection-union ratio is within a preset threshold range, wherein the threshold range is from a first threshold to a second threshold;

[0072] Step S74: when the intersection-union ratio is less than the first threshold, the output segment area has no band ends; when the intersection-union ratio is greater than the second threshold, the output segment area is in a transition stage; when the intersection-union ratio is between the first threshold and the second threshold, it is determined that the band end mask and the band side mask meet the preset conditions.

[0073] Specifically, the minimum x coordinate (left boundary), maximum x coordinate (right boundary), minimum y coordinate (upper boundary) and maximum y coordinate (lower boundary) of the end pixels in the tape end mask are recorded. The bounding box of the strip head is calculated based on the recorded coordinates. For the tape side mask, the bounding box of the strip side is obtained in the same way:

[0074] Compare the coordinates of the bounding box with ends and the bounding box with sides, determine the minimum x coordinate, maximum x coordinate, minimum y coordinate and maximum y coordinate of their overlapping parts to obtain their intersection bounding box; at the same time, compare the coordinates of the bounding box with ends and the bounding box with sides to determine their overall minimum x coordinate, maximum x coordinate, minimum y coordinate and maximum y coordinate to obtain their union bounding box. Calculate the area of ​​the intersection bounding box (width multiplied by height) and the area of ​​the union bounding box. Divide the area of ​​the intersection bounding box by the area of ​​the union bounding box to obtain the intersection-union ratio. When performing threshold range interval judgment, when the intersection-union ratio is less than the first threshold, the output segment area has no end, indicating that the segment area has no end. On the other hand, when the intersection-union ratio is greater than the second threshold, the output is in the transition stage.

[0075] It should be noted that in actual use, since the number of pixels of the intersection and union may be zero, exception handling is required before calculating the IOU to avoid division by zero errors.

[0076] Please continue reading Figure 7 , Figure 7The flowchart of step S71 in the method for locating the position of a steel strip provided in an embodiment of the present invention is as follows. In step S71, an end enclosing frame with an end mask and a side enclosing frame with a side mask are obtained respectively, including:

[0077] Step S711, performing morphological etching on the strip end mask and the strip side mask respectively to obtain a strip head fine mask and a strip side fine mask;

[0078] Step S712: Obtain the bounding box of the end portion and the bounding box of the side portion by calculating the connected domain bounding box for the fine mask of the steel head and the fine mask of the side portion.

[0079] It can be understood that for the morphological corrosion operation of the band end mask and the band side mask, a morphological operation method based on a structural element can be used, such as using an open operation or an corrosion operation. For example, the band end mask is morphologically eroded to shrink the detail area of ​​the band head. Then, the band end bounding box and the band side bounding box are obtained by using the above-mentioned connected domain analysis. Through morphological corrosion and connected domain analysis, the image can be effectively processed, the required target area can be extracted, and an accurate bounding box can be obtained, thereby improving the accuracy of recognition.

[0080] See also Figure 8 , Figure 8 The schematic diagram of the flow chart of step S80 in the strip steel location positioning method provided in the embodiment of the present invention. The positioning method further includes:

[0081] Step S81, counting the number of pixels with end masks;

[0082] Step S82, calculating the band end area according to the number of pixels of the band end mask;

[0083] Step S83, determining whether the area of ​​the belt end is within a preset area range, wherein the area range includes a first area threshold to a second area threshold;

[0084] Step S84: when the area of ​​the band end is smaller than the preset first area threshold, the output section area has no band end; when the area of ​​the band end is larger than the preset second area threshold, the output section area is in the transition stage; when the area of ​​the band end is between the first area threshold and the second area threshold, it is determined that the band end mask and the band side mask meet the preset conditions.

[0085] It can be understood that by counting the number of pixels in the band end mask, the overall area size of the band end can be understood. When the area of ​​the band end is small, the segment area is output as having no band end.

[0086] In step S80 , the area of ​​the bounding box with the ends may be directly used for determination.

[0087] In general, both step S70 and step S80 can judge various working conditions of the strip on the conveyor belt. Since the strip has a certain length in addition to its two ends (i.e., the strip head and the strip tail), in order to avoid misjudgment, step S70 uses the intersection and union ratio of the areas of the strip end bounding box and the strip side bounding box, and performs threshold judgment on the intersection and union ratio. Step S80 performs parallel judgment through the area of ​​the strip end mask. In actual conditions, only when one of the strip head and the strip side enter the strip image at the same time, or one of the strip tail and the strip side enter the strip image at the same time, will the execution instruction of starting the judgment section area at the strip head or the strip tail be determined; as mentioned above, since the strip end is very small, this method can greatly avoid misjudgment and improve the accuracy of recognition.

[0088] To further elaborate on the above, please refer to Fig. 9 , Fig. 9 Schematic diagram of the scenario that simply represents the strip image under various working conditions.

[0089] In the figure, a is a device for working with strip steel, b is a continuously moving conveyor belt, c is a steel belt placed on the conveyor belt, c1 represents the end of the steel belt, and the dotted area is the segment area where the strip steel image is collected.

[0090] Obviously, Fig. 9 (a) is a schematic diagram showing the situation when the strip head just enters the section area; Fig. 9 (b) is a schematic diagram showing the situation where the middle area of ​​the strip enters the segment area; Fig. 9 (c) is a schematic diagram showing the situation in the strip tail emergence area; Fig. 9 (d) is a schematic diagram representing a situation in which the tail of the strip with the current serial number and the head of the strip with the next serial number just enter the section area at the same time.

[0091] As the conveyor belt moves, the strip image gradually changes from Fig. 9 (a) Change to Fig. 9 (d) It can be seen that the change trend of the mask at the end of the tape can be achieved by running either step S40 or step S50 or both of them in parallel. Fig. 9 (a) and Fig. 9 (c) is the case where the strip tail or strip head is determined to avoid misjudgment at the stages shown in Figures (b) and (c) and improve the reliability of recognition.

[0092] See also Fig.10 , Fig.10 The flowchart of step S50 in the method for locating the position of the strip provided by the embodiment of the present invention is as follows. Threshold determination of the first difference to determine whether the segment area is the strip head or the strip tail includes:

[0093] Step S51, when the first difference is greater than 0, determining that the segment area is a strip head;

[0094] Step S52: when the first difference is less than 0, determine that this section of the area is the tail of the strip.

[0095] It can be understood that the first difference represents the difference between the mean values ​​of the horizontal coordinates of the mask with the end and the mask with the side. Figure 8 Understandably, Fig. 9 In case (a), the end mask is in front of the side mask. Fig. 9 In case (c), the end mask is behind the side mask, and in step 50, the direction along the pipeline is taken as a positive value, then the first difference is greater than 0, and in this case it is Fig. 9 (a) working condition, when the first difference is less than 0, Fig. 9 (c) working condition.

[0096] Please combine Figure 2 and Fig.11 , Fig.11 The schematic diagram of the flow chart of step S60 in the strip position positioning method provided in the embodiment of the present invention. Acquiring the position information corresponding to the strip head and the strip tail through the strip end mask includes:

[0097] Step S61, drawing the strip head position or the strip tail position according to the strip end bounding box, wherein the strip head position or the strip tail position both include an outer rectangle and a center point;

[0098] Step S62: output the strip head position or the strip tail position.

[0099] According to the previously determined strip end bounding box and the strip end type (i.e. strip head or strip tail) determined in step S50, the outer rectangle and center point corresponding to the position are drawn, specifically to obtain the bounding box information of the strip end. According to the coordinates of the upper left corner, width and height of the bounding box, the outer rectangle is drawn, the coordinates of the center point of the bounding box are calculated, and then the center point is drawn, and the above data is output.

[0100] Please combine Figure 2 and Figure 8 , Fig.12 , Fig.12 The flowchart of step S90 in the method for locating the position of a steel strip provided in an embodiment of the present invention is shown in FIG. The method for locating the position of a steel strip further includes:

[0101] Step S91, each time the position information of the strip head is obtained, the count is accumulated once to obtain a first value;

[0102] Step S92, calculating once each time the position information of the strip tail is obtained to obtain a second value;

[0103] Step S93: When the second difference between the first value and the second value is greater than 1 or less than 0, the output segment area is abnormal.

[0104] Understandably, according to Figure 2 According to the logical operation rules, when the strip steels with consecutive serial numbers are transported on the conveyor belt, taking the strip steel with serial number 1 as an example, the output achieved by this positioning method is as follows:

[0105] Serial number 1 → determine the section area as the strip head → output the position information of the strip head → the first value is accumulated to 1 → the output section area has no strip end → the output section area is the strip tail → output the position information of the strip head → the second value is accumulated to 1 → the output section area is the transition stage….

[0106] Then, from the strip with serial number 1 to the strip with serial number n, as long as it operates normally, the second difference between the first value and the second value will only be equal to 0 or 1. If the second difference is greater than 1 or less than 0, it proves that there is a problem with the strip head or strip tail of a certain serial number. At this time, an abnormality is prompted to screen out unqualified strips to avoid serious processing errors in subsequent processing.

[0107] Please combine Figure 2 and Figure 8 , Fig.13 , Fig.13 A schematic diagram of a module of a positioning system 100 for positioning a strip steel head is provided, and the positioning system 100 comprises:

[0108] An image acquisition module 10 is used to acquire a strip steel image in a segment area;

[0109] An extraction module 20 extracts a strip end mask and a strip side mask from the strip image based on semantic segmentation;

[0110] A first calculation module 30, when the band end mask and the band side mask meet a preset condition, calculates a first abscissa mean value and a second abscissa mean value of the band end mask and the band side mask respectively;

[0111] A second calculation module 40 is configured to obtain a first difference by subtracting the first horizontal coordinate mean from the second horizontal coordinate mean;

[0112] A judgment module 50, judging whether the segment area is a strip head or a strip tail according to the first difference;

[0113] The output module 60 obtains the position information corresponding to the strip head and the strip tail through the strip end mask.

[0114] The positioning system 100 can be packaged in a device to achieve the same effect as the above method, which will not be repeated here.

[0115] It should be understood that the specific features, operations and details described hereinabove with respect to the method of the present application may also be similarly applied to the device and system of the present application, or vice versa. In addition, each step of the method of the present application described above may be performed by a corresponding component or unit of the device or system of the present application.

[0116] It should be understood that each module / unit of the device of the present application can be implemented in whole or in part by software, hardware, firmware or a combination thereof. Each module / unit can be embedded in the processor of the electronic device or independent of the processor in the form of hardware or firmware, or can be stored in the memory of the electronic device in the form of software for the processor to call to perform the operation of each module / unit. Each module / unit can be implemented as an independent component or module, or two or more modules / units can be implemented as a single component or module.

[0117] The present application also provides an electronic device, the electronic device comprising a processor and a memory storing computer program instructions. When the processor executes the computer program instructions, each step of the above-mentioned strip position positioning method is implemented. The electronic device can be broadly a server, a terminal, or any other electronic device with necessary computing and / or processing capabilities.

[0118] In one embodiment, the electronic device may include a processor, a memory, a network interface, a communication interface, etc. connected through a system bus. The processor of the electronic device may be used to provide necessary computing, processing and / or control capabilities. The memory of the electronic device may include a non-volatile storage medium and an internal memory. The non-volatile storage medium may store an operating system, a computer program, etc. The internal memory may provide an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface and the communication interface of the electronic device may be used to connect and communicate with external devices through a network. When the computer program is executed by the processor, the steps of the method of the present application are executed.

[0119] The present application provides a computer-readable storage medium, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the above-mentioned strip position positioning method is implemented.

[0120] Those skilled in the art will appreciate that the method steps of the present application can be completed by instructing related hardware such as electronic devices or processors through a computer program, and the computer program can be stored in a non-temporary computer-readable storage medium, and the computer program causes the steps of the present application to be executed when it is executed. Depending on the circumstances, any reference to memory, storage or other media herein may include non-volatile or volatile memory. Examples of non-volatile memory include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, magnetic tape, floppy disk, magneto-optical data storage device, optical data storage device, hard disk, solid state disk, etc. Examples of volatile memory include random access memory (RAM), external cache memory, etc.

[0121] The various technical features described above can be combined arbitrarily. Although all possible combinations of these technical features are not described, any combination of these technical features should be considered to be covered by this specification as long as there is no contradiction in such combination.

[0122] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for locating a strip steel position, characterized in that: include: Acquire a strip image of a segment area; Extracting a strip end mask and a strip side mask in the strip steel image based on semantic segmentation; When the band end mask and the band side mask meet a preset condition, respectively calculating a first abscissa mean value and a second abscissa mean value of the band end mask and the band side mask; Subtracting the first horizontal coordinate mean from the second horizontal coordinate mean to obtain a first difference; Performing a threshold judgment on the first difference to determine whether the segment area is a strip head or a strip tail; The position information corresponding to the strip head and the strip tail is obtained through the strip end mask.

2. The positioning method according to claim 1, characterized in that: The extracting of the strip end mask and the strip side mask in the strip steel image based on semantic segmentation includes: Define the category labels of end, side, and background respectively; Performing semantic segmentation on the steel strip image according to the category label using a pre-trained semantic segmentation model, and extracting a mask image in the prediction result; An end mask and a side mask are obtained from the mask image.

3. The positioning method according to claim 1, characterized in that: The first abscissa mean and the second abscissa mean according to the tape end mask and the tape side mask include: Traverse each pixel with an end mask and a pixel with a side mask respectively, and record all first position coordinates and second position coordinates; Add the horizontal coordinates of the first position coordinates and divide the sum by the number of pixels of the first position coordinates to obtain a first horizontal coordinate mean; The horizontal coordinates of the second position coordinates are added and divided by the number of pixels of the second position coordinates to obtain a second horizontal position coordinate mean.

4. The positioning method according to claim 1, characterized in that: After extracting the strip end mask and the strip side mask in the strip steel image based on semantic segmentation, the method further includes: Obtain an end bounding box with an end mask and a side bounding box with a side mask respectively; Calculate the ratio of the intersection and union of the bounding box with the ends and the bounding box with the sides; Determine whether the intersection-union ratio is within a preset threshold range, where the threshold range is from a first threshold to a second threshold; When the intersection-union ratio is less than the first threshold, the segment area is outputted as having no band ends; when the intersection-union ratio is greater than the second threshold, the segment area is outputted as being in a transition stage; when the intersection-union ratio is between the first threshold and the second threshold, it is determined that the band end mask and the band side mask meet preset conditions.

5. The positioning method according to claim 4, characterized in that: The step of respectively obtaining an end bounding box with an end mask and a side bounding box with a side mask comprises: The strip end mask and the strip side mask are morphologically etched respectively to obtain the strip head fine mask and the strip side fine mask; The bounding box with ends and the bounding box with sides are obtained by calculating the connected domain bounding box for the fine mask with steel head and the fine mask with sides.

6. The positioning method according to claim 4, characterized in that: After extracting the strip end mask and the strip side mask in the strip steel image based on semantic segmentation, the method further includes: Counting the number of pixels with end masks; Calculate the area of ​​the band end according to the number of pixels of the band end mask; Determining whether the area of ​​the belt end is within a preset area range, wherein the area range includes a first area threshold to a second area threshold; When the area of ​​the band end is smaller than a preset first area threshold, it is output that the segment area has no band end; when the area of ​​the band end is larger than a preset second area threshold, it is output that the segment area is in a transition stage; when the area of ​​the band end is between the first area threshold and the second area threshold, it is determined that the band end mask and the band side mask meet the preset conditions.

7. The positioning method according to claim 4, characterized in that: The obtaining of the position information corresponding to the strip head and the strip tail through the strip end mask comprises: Draw the position of the strip head according to the strip end bounding box, wherein the strip head position includes an outer rectangle and a center point; The strip head position is output.

8. The positioning method according to claim 4, characterized in that: The step of judging whether the segment area is a strip head or a strip tail according to the first difference value comprises: When the first difference is greater than 0, determining that the segment area is a strip steel head; When the first difference is less than 0, it is determined that this section of the area is the tail of the strip.

9. The positioning method according to any one of claims 1 to 8, characterized in that: The positioning method further comprises: Each time the position information of the strip steel head is obtained, the number is accumulated once to obtain a first value; Each time the position information of the strip tail is obtained, the second value is calculated once to obtain the second value; When the second difference between the first value and the second value is greater than 1 or less than 0, the segment area is output as abnormal.

10. A positioning system for a strip steel head, characterized in that: include: An image acquisition module, which acquires the strip steel image of the segment area; An extraction module extracts a strip end mask and a strip side mask from the strip steel image based on semantic segmentation; A first calculation module, which calculates a first abscissa mean and a second abscissa mean of the band end mask and the band side mask respectively when the band end mask and the band side mask meet a preset condition; A second calculation module is configured to obtain a first difference by subtracting the first horizontal coordinate mean from the second horizontal coordinate mean; A judging module, for judging whether the segment area is a strip head or a strip tail according to the first difference; The output module obtains the position information corresponding to the strip head and the strip tail through the strip end mask.

Images