Method, device and equipment for detecting surface defects of hot-rolled strip steel and storage medium

Abstract

The present application relates to the technical field of strip steel defect detection, and specifically discloses a hot-rolled strip steel surface defect detection method, device, equipment and storage medium, comprising: in the process of surface defect detection of strip steel on a roller bed by an industrial line scanning camera, offset analysis is performed on the strip steel on the roller bed, the offset degree of the strip steel on the roller bed is evaluated, the offset situation of the strip steel in distance and angle is determined, after accurate coordinate correction of the collected image, the actual position of the defect on the strip steel surface can be improved during surface defect detection, the defect position judgment error caused by offset is avoided, and the tension control of the roller bed can also be adjusted according to the offset situation of the strip steel, so that the interference degree of the surface defect detection of the strip steel caused by the offset of the strip steel is reduced while reducing the wear of the roller bed.

Description

Technical Field

[0001] This invention relates to the field of strip steel defect detection technology, specifically to methods, apparatus, equipment, and storage media for detecting surface defects in hot-rolled strip steel. Background Technology

[0002] In the steel industry, hot-rolled strip steel is an important basic material, and its surface quality directly affects the performance and quality of subsequent processed products. During the production process of hot-rolled strip steel, factors such as complex rolling technology, high rolling speed, large temperature variations of the strip steel, and unstable roller table operation can easily lead to strip deviation on the roller table.

[0003] Currently, industrial line scanning cameras are widely used in industry for defect detection on the surface of hot-rolled strip steel. Line scanning cameras, with their high resolution and high speed, can acquire image information of the strip steel surface in real time, providing data support for subsequent defect identification and analysis. However, in actual production scenarios, when the strip steel deviates significantly on the roller conveyor, this deviation causes a discrepancy between the image acquired by the line scanning camera and the actual surface position of the strip steel. This deviation makes it difficult to accurately determine the actual location of defects on the strip steel surface during subsequent defect detection, easily leading to incorrect defect location judgments. Furthermore, severe strip steel deviation can cause parts of the strip steel surface to exceed the field of view of the line scanning camera, forming undetected areas. The existence of these undetected areas prevents the defect detection system from comprehensively inspecting the strip steel surface, thus missing some potential defects. More seriously, because the degree of interference from undetected areas on the defect detection process cannot be accurately measured, it is difficult for inspectors to judge the reliability of the detection results.

[0004] Furthermore, there are various types of missed detection zones in the field of view, mainly categorized into angular offset and distance offset zones. Different types of missed detection zones have different impact mechanisms on defect detection; without accurate differentiation and analysis, effective corrective measures cannot be taken for different types of missed detection problems.

[0005] Therefore, the present invention provides a method, apparatus, equipment and storage medium for detecting surface defects in hot-rolled strip steel. Summary of the Invention

[0006] The purpose of this invention is to provide a method, apparatus, equipment, and storage medium for detecting surface defects in hot-rolled strip steel, in order to solve the aforementioned background problems.

[0007] The objective of this invention can be achieved through the following technical solutions: Methods for detecting surface defects in hot-rolled strip steel include: During the process of surface defect detection of strip steel on roller conveyor using an industrial line scan camera, the offset of strip steel on roller conveyor is analyzed to evaluate the degree of offset of strip steel on roller conveyor. In scenarios where strip steel is severely misaligned on the roller conveyor, the undetected area on the surface of the strip steel on the roller conveyor is obtained, and the undetected area is analyzed to determine the interference degree of strip steel misalignment. An analysis of the out-of-view type was conducted on the undetected areas on the surface of the strip on the roller conveyor, and out-of-view areas with angular offset and distance offset were screened out. For the selected angle offset and distance offset areas outside the field of view, determine the over-angle offset correction amount and the over-distance offset correction amount respectively, and perform correction operations on the missed areas of the field of view based on the over-angle offset correction amount and the over-distance offset correction amount.

[0008] As a further aspect of the present invention, the process of analyzing the offset of the strip steel on the roller conveyor is as follows: The strip surface is divided into several surface detection sub-regions according to the grid method. The strip offset monitoring cycle is set and the strip offset monitoring cycle is equally divided into several strip offset monitoring time periods. In each strip offset monitoring time period, the actual center line in each surface detection sub-region and the nominal center line calibrated by the line scan camera in the surface detection sub-region are extracted. Obtain the distance between the actual centerline and the nominal centerline, and calculate the ratio between the actual centerline length and the nominal centerline length as the sub-region offset distance ratio. The average periodic offset distance is obtained by summing and averaging the sub-region offset distance ratios within all surface detection sub-regions. During each strip offset detection period, the angle between the actual centerline and the nominal centerline in each surface detection sub-region is obtained, and the ratio of this angle to the flat angle is calculated to obtain the sub-region offset angle ratio. The average periodic offset angle is obtained by summing and averaging the sub-region offset angle ratios within all surface detection sub-regions.

[0009] As a further aspect of the present invention, the evaluation process for the degree of strip deflection on the roller conveyor is as follows; The average value of the periodic offset distance and the average value of the periodic offset angle are summed to obtain the strip offset detection value. If the strip offset detection value is greater than the strip offset detection threshold, it is displayed as a severe strip offset signal.

[0010] A further aspect of the present invention is as follows: the process of obtaining the undetected areas on the surface of the strip on the roller conveyor and analyzing the undetected areas is as follows: The surface monitoring sub-region image is compared with the surface monitoring sub-region on the strip surface. The surface monitoring sub-region image that does not overlap with the surface monitoring sub-region on the strip surface is marked as the field of view missed detection area. The proportion of the number of field of view missed detection areas to the total number of surface monitoring sub-regions on the strip surface is counted as the field of view missed detection ratio. The area difference between the surface monitoring sub-region area on the strip surface and the surface monitoring sub-region image is obtained, and the ratio of the surface monitoring sub-region area is calculated to obtain the field of view missed detection ratio. The average degree of field-of-view misses is obtained by summing and averaging the ratios of the degree of field-of-view misses in all missed areas.

[0011] As a further aspect of the present invention, the process for determining the interference degree of strip offset detection is as follows: The offset missed detection interference value is calculated by multiplying the ratio of the number of missed detections in the field of view with the average degree of missed detections in the field of view. If the offset missed detection interference value is greater than the offset missed detection interference threshold, it is displayed as an offset height missed detection signal.

[0012] As a further aspect of the present invention, the process of performing over-the-horizon type analysis on the undetected areas on the surface of the strip on the roller conveyor is as follows: Extract the actual centerline within the undetected area of ​​the field of view and compare it with the nominal centerline calibrated by the line scan camera. If there is no angle between the actual centerline and the nominal centerline within the undetected area of ​​the field of view and the two are in a parallel position, then the undetected area of ​​the field of view is marked as a distance offset beyond the field of view. If there is an angle between the actual centerline and the nominal centerline within the analyzed field of view's missed detection area, and their positional relationship is intersecting, then the analyzed field of view's missed detection area is marked as an angle-shifted out-of-view area.

[0013] As a further aspect of the present invention, the process of determining the over-angle deviation correction amount and the over-range deviation correction amount, and performing a correction operation on the missed detection area of ​​the field of view is as follows: For the angle deviation beyond the field of view, the angle between the actual center line and the nominal center line within the angle deviation beyond the field of view is extracted as the angle deviation correction amount. Based on the angle deviation correction amount, the surface monitoring sub-region image corresponding to the angle deviation beyond the field of view is rotated and corrected. For distance offset beyond the line of sight, the distance between the actual centerline and the nominal centerline within the distance offset beyond the line of sight is extracted as the distance offset correction amount. The surface monitoring sub-area image corresponding to the distance offset beyond the line of sight is then translated and corrected based on the distance offset correction amount by physical correction methods such as adjusting roller tension, adjusting side guide roller opening / pressure, and synchronous roller linear speed.

[0014] A surface defect detection device for hot-rolled strip steel includes: Strip offset assessment module: During the process of surface defect detection of strip steel on the roller table using an industrial line scan camera, the offset of the strip steel on the roller table is analyzed to assess the degree of strip offset. Offset Missed Detection Interference Module: In scenarios where the strip steel is severely offset on the roller conveyor, the module acquires the field of view of the missed detection area on the surface of the strip steel on the roller conveyor, analyzes the field of view of the missed detection area, and determines the degree of strip steel offset missed detection interference. The field of view omission category module analyzes the out-of-view type of the out-of-view area on the surface of the strip on the roller conveyor, and filters out the out-of-view areas with angular deviation and distance deviation. Field of view omission correction module: For the selected angle offset and distance offset out-of-view areas, determine the over-angle offset correction amount and over-distance offset correction amount respectively, and perform correction operation on the field of view omission area based on the over-angle offset correction amount and over-distance offset correction amount.

[0015] Hot-rolled strip steel surface defect detection equipment includes: The method includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the hot-rolled strip surface defect detection method as described in any one of claims 1-7.

[0016] Storage medium for detecting surface defects in hot-rolled strip steel includes: The computer program causes the computer to execute the hot-rolled strip surface defect detection method as described in any one of claims 1-7.

[0017] The beneficial effects of this invention are as follows: 1. This invention utilizes an industrial line scan camera to detect surface defects in strip steel on a roller conveyor. It analyzes the offset of the strip steel on the roller conveyor, assesses the degree of offset, and clarifies the offset in distance and angle. After precise coordinate correction of the acquired images, it can improve the actual location of defects on the strip steel surface during surface defect detection, avoiding errors in defect location judgment due to offset. Furthermore, it can adjust the tension control of the roller conveyor based on the strip steel offset, reducing roller conveyor wear and minimizing interference with surface defect detection caused by strip steel offset.

[0018] 2. In scenarios where strip steel is severely misaligned on the roller conveyor, this invention acquires the undetected areas on the surface of the strip steel on the roller conveyor and analyzes these undetected areas. This allows for a comprehensive assessment of the interference of undetected areas caused by strip steel misalignment on the surface defect detection process, providing an important basis for judging the reliability of defect detection results. By analyzing the out-of-view types of undetected areas on the surface of the strip steel on the roller conveyor, angular misalignment out-of-view areas and distance misalignment out-of-view areas are identified. This allows inspection personnel to clearly understand the specific types of undetected areas caused by strip steel misalignment, which helps to more accurately locate the actual position of defects on the strip steel. Moreover, the occurrence of angular misalignment out-of-view areas and distance misalignment out-of-view areas is often related to the state of equipment such as roller conveyors and transmission devices, and can also serve as a basis for evaluating the stability of equipment operation. Attached Figure Description

[0019] The invention will now be further described with reference to the accompanying drawings.

[0020] Figure 1 This is a functional module diagram of the hot-rolled strip steel surface defect detection method of the present invention; Figure 2 This is a flowchart illustrating the determination process of the hot-rolled strip steel surface defect detection method in this invention. Figure 3 This is a flowchart of the modules within the hot-rolled strip steel surface defect detection device of the present invention. Detailed Implementation

[0021] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0022] Example 1 like Figure 1 - Figure 2 As shown, this embodiment of the invention provides a method for detecting surface defects in hot-rolled strip steel, including: Step 1: During the surface defect detection of strip steel on the roller conveyor using an industrial line scan camera, the offset of the strip steel on the roller conveyor is analyzed to evaluate the degree of offset of the strip steel on the roller conveyor. In some embodiments, a strip offset monitoring cycle is set, and the strip offset monitoring cycle is equally divided into several strip offset monitoring periods, wherein the duration of each strip offset monitoring period is equal. The surface of the strip steel is divided into several surface inspection sub-regions according to a grid pattern, wherein each surface inspection sub-region occupies an equal area of ​​the strip steel surface; It should be noted that one surface detection sub-area is monitored within each strip offset detection period; During each strip offset detection period, the actual center line in each surface detection sub-area and the nominal center line calibrated by the line scan camera in the surface detection sub-area are extracted. Obtain the distance between the actual centerline and the nominal centerline, and calculate the ratio between the actual centerline length and the nominal centerline length as the sub-region offset distance ratio. The average periodic offset distance is obtained by summing and averaging the sub-region offset distance ratios within all surface detection sub-regions. Similarly, within each strip offset detection period, the angle between the actual centerline and the nominal centerline in each surface detection sub-region is obtained and compared with the flat angle (180°). 0 The ratio is calculated to obtain the sub-region offset angle ratio; It should be noted that when the actual centerline within the surface detection sub-area and the nominal centerline calibrated by the line scanner within the surface detection sub-area do not deviate, they coincide, and the angle formed between them is a flat angle (180°).0 ); The average periodic offset angle is obtained by summing and averaging the offset angle ratios of all sub-regions within the surface detection sub-regions. The average value of the periodic offset distance and the average value of the periodic offset angle are summed to obtain the strip offset detection value; It is understandable that the strip offset detection value represents the following meaning: it comprehensively reflects the strip's offset in two key dimensions, distance and angle. On the one hand, the average periodic offset distance reflects the average level of offset in the distance direction of each sub-region on the strip surface throughout the entire monitoring period, reflecting the overall offset trend of the strip in the distance dimension. On the other hand, the average periodic offset angle reflects the average level of offset in the angle direction of each sub-region on the strip surface throughout the entire monitoring period, reflecting the overall offset trend of the strip in the angle dimension. Specifically, if the strip offset detection value is larger, it indicates that the overall offset of the strip is more serious during the monitoring period, including positional offset in the horizontal or vertical direction (distance dimension) and offset in rotation or tilt (angle dimension). If the strip offset detection value is smaller, it indicates that the overall offset of the strip is less serious during the monitoring period, including positional offset in the horizontal or vertical direction (distance dimension) and offset in rotation or tilt (angle dimension). If the strip offset detection value is greater than the strip offset detection threshold, it indicates that the overall offset of the strip is relatively serious during the monitoring period, which is displayed as a serious strip offset signal. If the strip offset detection value is less than or equal to the strip offset detection threshold, it indicates that the overall offset of the strip is relatively slight during the monitoring period, which is displayed as a slight strip offset signal. The specific solution in this embodiment is as follows: During the surface defect detection of strip steel on the roller conveyor using an industrial line scanning camera, the offset of the strip steel on the roller conveyor is analyzed to assess the degree of offset of the strip steel on the roller conveyor, and to clarify the offset of the strip steel in terms of distance and angle. After accurate coordinate correction of the acquired image, the actual position of the defect on the surface of the strip steel can be improved during surface defect detection, avoiding errors in defect position judgment caused by offset. Moreover, the tension control of the roller conveyor can be adjusted according to the offset of the strip steel, reducing roller conveyor wear and reducing the interference of strip steel offset on the detection of surface defects of the strip steel.

[0023] Example 2 like Figure 1 - Figure 2 As shown, this embodiment of the invention provides a method for detecting surface defects in hot-rolled strip steel, including: Step 2: In the scenario where the strip steel is severely misaligned on the roller conveyor, obtain the undetected area on the surface of the strip steel on the roller conveyor, and analyze the undetected area to determine the interference degree of strip steel misalignment. In some embodiments, surface monitoring sub-region images captured by an industrial line scan camera during each strip offset monitoring period are extracted; The surface monitoring sub-region image is compared with the surface monitoring sub-region on the strip surface. The surface monitoring sub-region image that does not overlap with the surface monitoring sub-region on the strip surface is marked as the field of view missed area. The surface monitoring sub-region image that overlaps with the surface monitoring sub-region on the strip surface is marked as a non-field-of-view missed detection area; For example, the process of performing region overlap comparison between the surface monitoring sub-region image and the surface monitoring sub-region on the strip surface is as follows: If the area of ​​the surface monitoring sub-region image is inconsistent with the area of ​​the surface monitoring sub-region on the strip surface, it is marked as a missed detection area in the field of view; It is understandable that when the surface monitoring sub-region image is smaller than the surface monitoring sub-region area on the strip surface, that is, the local area of ​​the surface monitoring sub-region on the strip surface exceeds the surface monitoring area captured by the industrial line scan camera; If the area of ​​the surface monitoring sub-region image is the same as the area of ​​the surface monitoring sub-region on the strip surface, it is marked as a non-field-of-view missed detection area; Understandably, although the surface monitoring sub-region appears to be offset in the surface monitoring sub-region image, the surface monitoring sub-region on the strip surface does not exceed the surface monitoring area captured by the industrial line scan camera. The proportion of the number of undetected areas in the field of view to the total number of surface monitoring sub-regions on the strip surface is used as the undetected area ratio. The area difference between the surface monitoring sub-region area on the strip surface and the surface monitoring sub-region image is obtained, and the ratio of the surface monitoring sub-region area is calculated to obtain the field of view missed detection ratio. The average degree of field-of-view misses is obtained by summing and averaging the ratios of the degree of field-of-view misses in all the missed areas. The offset missing detection interference value is obtained by multiplying the ratio of the number of missed detections in the field of view with the average degree of missed detections in the field of view. It is understandable that the offset omission interference value represents the degree of interference of the field of view omission caused by the strip offset on the surface defect detection process in the scenario of severe strip offset on the roller conveyor. It is considered from two key aspects: the number of omission areas and the degree of omission. On the one hand, the ratio of the number of omissions in the field of view reflects the overall situation that the camera failed to capture the surface monitoring sub-area completely due to the strip offset. On the other hand, the average degree of omission in the field of view reflects the average level of omission in all omission areas. Specifically, if the offset omission interference value is larger, it means that there are more surface monitoring sub-areas that are not completely captured by the camera due to the strip offset, and the proportion of the area that is not completely captured by the camera is larger, that is, the degree of omission is larger. If the offset omission interference value is smaller, it means that there are fewer surface monitoring sub-areas that are not completely captured by the camera due to the strip offset, and the proportion of the area that is not completely captured by the camera is smaller, that is, the degree of omission is smaller. If the offset missed detection interference value is greater than the offset missed detection interference threshold, it indicates that there are a large number of surface monitoring sub-areas that are not fully captured by the camera due to strip offset, and the proportion of areas not fully captured by the camera is large, that is, the degree of missed detection is large, which is displayed as an offset height missed detection signal. If the offset missed detection interference value is less than or equal to the offset missed detection interference threshold, it indicates that the number of surface monitoring sub-areas that were not fully captured by the camera due to strip offset is small, and the proportion of areas not fully captured by the camera is small, that is, the degree of missed detection is small, and it is displayed as a low-degree offset missed detection signal.

[0024] Step 3: Analyze the out-of-view type of the undetected areas on the surface of the strip on the roller conveyor, and screen out the out-of-view areas with angular offset and distance offset. In some embodiments, the actual center line within the field of view missing detection area is extracted and compared with the nominal center line calibrated by the line scan camera. If there is no angle between the actual center line and the nominal center line within the analyzed field of view missing detection area, and the positional relationship between the two is parallel, then the analyzed field of view missing detection area is marked as a distance offset beyond the field of view. If there is an angle between the actual centerline and the nominal centerline within the analyzed field of view's missed detection area, and their positional relationship is intersecting, then the analyzed field of view's missed detection area is marked as an angle-shifted out-of-view area. It should be noted that the significance of screening out angular offset and distance offset from the line of sight is that it allows inspectors to clearly understand the specific type of missed detection caused by strip offset. For angular offset from the line of sight, it can be determined that the strip has rotated or tilted, causing an angular deviation between the camera's shooting range and the actual strip area. For distance offset from the line of sight, it indicates that the strip has shifted horizontally or vertically, causing some areas to exceed the camera's shooting range. This clear identification of the type of positional deviation helps to more accurately locate the actual position of defects on the strip. Meanwhile, since the occurrence of angular deviation and distance deviation beyond the line of sight is often related to the condition of equipment such as roller conveyors and transmission devices, when angular deviation beyond the line of sight occurs, it indicates that there are problems such as uneven rotation of the roller conveyor or loose transmission chain; when distance deviation beyond the line of sight occurs, it indicates that the horizontal or vertical positioning of the roller conveyor is inaccurate or the support structure is loose. Therefore, it can also serve as a basis for evaluating the stability of equipment operation. The specific solution in this embodiment is as follows: In the scenario of severe strip misalignment on the roller conveyor, the undetected area of ​​the strip surface on the roller conveyor is obtained, and the undetected area is analyzed. This can comprehensively measure the degree of interference of the undetected area caused by the strip misalignment on the surface defect detection process, and provide an important basis for judging the reliability of the defect detection results. By performing an over-the-view type analysis on the undetected area of ​​the strip surface on the roller conveyor, the angular offset over-the-view area and the distance offset over-the-view area are screened out. This allows the inspection personnel to clearly understand the specific type of undetected area caused by the strip misalignment, which helps to more accurately locate the actual position of the defect on the strip. Moreover, the occurrence of angular offset over-the-view area and distance offset over-the-view area is often related to the state of equipment such as roller conveyors and transmission devices, and can also serve as a basis for evaluating the stability of equipment operation.

[0025] Example 3 like Figure 1 - Figure 2 As shown, this embodiment of the invention provides a method for detecting surface defects in hot-rolled strip steel, which further includes: Step 4: For the selected angle offset and distance offset areas outside the field of view, determine the over-angle offset correction amount and the over-distance offset correction amount respectively, and perform correction operations on the missed areas of the field of view based on the over-angle offset correction amount and the over-distance offset correction amount. For example, for angle deviation beyond the field of view, the angle between the actual centerline and the nominal centerline within the angle deviation beyond the field of view is extracted as the angle deviation correction amount. The process of correcting the angular deviation beyond the field of view based on the over-viewing angle correction amount is as follows: Based on the over-viewing angle offset correction amount, the surface monitoring sub-region image corresponding to the angle offset beyond the viewing area is rotated and corrected. For distance offset beyond the line of sight, the distance between the actual centerline and the nominal centerline within the distance offset beyond the line of sight is extracted as the distance offset correction amount beyond the line of sight. The process of correcting distance offset beyond the line of sight based on the beyond-line-of-sight offset correction is as follows: The surface monitoring sub-area image corresponding to the distance offset beyond the line of sight is translated and corrected by adjusting the roller tension, adjusting the opening degree / pressure of the side guide roller, and adjusting the linear speed of the synchronous roller, based on the distance offset correction amount.

[0026] Example 4 Please see Figure 3 As shown, the present invention also provides a surface defect detection device for hot-rolled strip steel, comprising the following modules: Strip offset assessment module: During the process of surface defect detection of strip steel on the roller table using an industrial line scan camera, the offset of the strip steel on the roller table is analyzed to assess the degree of strip offset. Offset Missed Detection Interference Module: In scenarios where the strip steel is severely offset on the roller conveyor, the module acquires the field of view of the missed detection area on the surface of the strip steel on the roller conveyor, analyzes the field of view of the missed detection area, and determines the degree of strip steel offset missed detection interference. The field of view omission category module analyzes the out-of-view type of the out-of-view area on the surface of the strip on the roller conveyor, and filters out the out-of-view areas with angular deviation and distance deviation. Field of view omission correction module: For the selected angle offset and distance offset out-of-view areas, determine the over-angle offset correction amount and over-distance offset correction amount respectively, and perform correction operation on the field of view omission area based on the over-angle offset correction amount and over-distance offset correction amount.

[0027] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the present invention should still fall within the scope of the present invention.

Claims

1. A method for detecting surface defects in hot-rolled strip steel, characterized in that: include: During the process of surface defect detection of strip steel on roller conveyor using an industrial line scan camera, the offset of strip steel on roller conveyor is analyzed to evaluate the degree of offset of strip steel on roller conveyor. In scenarios where strip steel is severely misaligned on the roller conveyor, the undetected area on the surface of the strip steel on the roller conveyor is obtained, and the undetected area is analyzed to determine the interference degree of strip steel misalignment. An analysis of the out-of-view type was conducted on the undetected areas on the surface of the strip on the roller conveyor, and out-of-view areas with angular offset and distance offset were screened out. For the selected angle offset and distance offset areas outside the field of view, determine the over-angle offset correction amount and the over-distance offset correction amount respectively, and perform correction operations on the missed areas of the field of view based on the over-angle offset correction amount and the over-distance offset correction amount.

2. The method for detecting surface defects in hot-rolled strip steel according to claim 1, characterized in that: The process of analyzing the offset of the strip steel on the roller conveyor is as follows: The strip surface is divided into several surface detection sub-regions according to the grid method. The strip offset monitoring cycle is set and the strip offset monitoring cycle is equally divided into several strip offset monitoring time periods. In each strip offset monitoring time period, the actual center line in each surface detection sub-region and the nominal center line calibrated by the line scan camera in the surface detection sub-region are extracted. Obtain the distance between the actual centerline and the nominal centerline, and calculate the ratio between the actual centerline length and the nominal centerline length as the sub-region offset distance ratio. The average periodic offset distance is obtained by summing and averaging the sub-region offset distance ratios within all surface detection sub-regions. During each strip offset detection period, the angle between the actual centerline and the nominal centerline in each surface detection sub-region is obtained, and the ratio of this angle to the flat angle is calculated to obtain the sub-region offset angle ratio. The average periodic offset angle is obtained by summing and averaging the sub-region offset angle ratios within all surface detection sub-regions.

3. The method for detecting surface defects in hot-rolled strip steel according to claim 2, characterized in that: The evaluation process for the degree of strip deflection on the roller conveyor is as follows; The average value of the periodic offset distance and the average value of the periodic offset angle are summed to obtain the strip offset detection value. If the strip offset detection value is greater than the strip offset detection threshold, it is displayed as a severe strip offset signal.

4. The method for detecting surface defects in hot-rolled strip steel according to claim 1, characterized in that: The process of obtaining the undetected areas on the surface of the strip on the roller conveyor and analyzing these undetected areas is as follows: The surface monitoring sub-region image is compared with the surface monitoring sub-region on the strip surface. The surface monitoring sub-region image that does not overlap with the surface monitoring sub-region on the strip surface is marked as the field of view missed detection area. The proportion of the number of field of view missed detection areas to the total number of surface monitoring sub-regions on the strip surface is counted as the field of view missed detection ratio. The area difference between the surface monitoring sub-region area on the strip surface and the surface monitoring sub-region image is obtained, and the ratio of the surface monitoring sub-region area is calculated to obtain the field of view missed detection ratio. The average degree of field-of-view misses is obtained by summing and averaging the ratios of the degree of field-of-view misses in all missed areas.

5. The method for detecting surface defects in hot-rolled strip steel according to claim 4, characterized in that: The process for determining the interference degree of strip offset missed detection is as follows: The offset missed detection interference value is calculated by multiplying the ratio of the number of missed detections in the field of view with the average degree of missed detections in the field of view. If the offset missed detection interference value is greater than the offset missed detection interference threshold, it is displayed as an offset height missed detection signal.

6. The method for detecting surface defects in hot-rolled strip steel according to claim 4, characterized in that: The process of performing over-the-horizon type analysis on the undetected areas on the surface of the strip on the roller conveyor is as follows: Extract the actual centerline within the undetected area of ​​the field of view and compare it with the nominal centerline calibrated by the line scan camera. If there is no angle between the actual centerline and the nominal centerline within the undetected area of ​​the field of view and the two are in a parallel position, then the undetected area of ​​the field of view is marked as a distance offset beyond the field of view. If there is an angle between the actual centerline and the nominal centerline within the analyzed field of view's missed detection area, and their positional relationship is intersecting, then the analyzed field of view's missed detection area is marked as an angle-shifted out-of-view area.

7. The method for detecting surface defects in hot-rolled strip steel according to claim 1, characterized in that: The process of determining the over-angle deviation correction and over-range deviation correction, and then correcting the missed areas in the field of view is as follows: For the angle deviation beyond the field of view, the angle between the actual center line and the nominal center line within the angle deviation beyond the field of view is extracted as the angle deviation correction amount. Based on the angle deviation correction amount, the surface monitoring sub-region image corresponding to the angle deviation beyond the field of view is rotated and corrected. For distance offset beyond the line of sight, the distance between the actual centerline and the nominal centerline within the distance offset beyond the line of sight is extracted as the distance offset correction amount. The surface monitoring sub-area image corresponding to the distance offset beyond the line of sight is then translated and corrected based on the distance offset correction amount by physical correction methods such as adjusting roller tension, adjusting side guide roller opening / pressure, and synchronous roller linear speed.

8. A surface defect detection device for hot-rolled strip steel, implementing the surface defect detection method for hot-rolled strip steel as described in any one of claims 1-7, characterized in that: Includes the following modules: Strip offset assessment module: During the process of surface defect detection of strip steel on the roller table using an industrial line scan camera, the offset of the strip steel on the roller table is analyzed to assess the degree of strip offset. Offset Missed Detection Interference Module: In scenarios where the strip steel is severely offset on the roller conveyor, the module acquires the field of view of the missed detection area on the surface of the strip steel on the roller conveyor, analyzes the field of view of the missed detection area, and determines the degree of strip steel offset missed detection interference. The field of view omission category module analyzes the out-of-view type of the out-of-view area on the surface of the strip on the roller conveyor, and filters out the out-of-view areas with angular deviation and distance deviation. Field of view omission correction module: For the selected angle offset and distance offset out-of-view areas, determine the over-angle offset correction amount and over-distance offset correction amount respectively, and perform correction operation on the field of view omission area based on the over-angle offset correction amount and over-distance offset correction amount.

9. A surface defect detection device for hot-rolled strip steel, characterized in that: The method includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the hot-rolled strip surface defect detection method as described in any one of claims 1-7.

10. A storage medium for detecting surface defects in hot-rolled strip steel, containing a computer program, characterized in that: The computer program causes the computer to execute the hot-rolled strip surface defect detection method as described in any one of claims 1-7.

Images