Anomaly detection device and anomaly detection method

Abstract

An abnormality detection device and an abnormality detection method are provided that are capable of detecting rolling abnormalities in metal strips with higher accuracy than conventional devices. [Solution] An image processing unit (92) sets a detection area (120) of the metal strip (1) to be detected for abnormalities from the latest image, and also sets a reflected light area (101) and an outlying reflected light area (103) reflected on the surface of the metal strip (1) that are caused by the illumination light illuminating the metal strip (1). The area obtained by removing the reflected light area (101) and the outlying reflected light area (103) from the detection area (120) is set as a new detection area, and abnormal areas (111, 113) are set based on the relationship between the brightness value difference between the brightness data of the pixels in the new detection area and the brightness reference value, thereby detecting rolling abnormalities.

Description

[Technical Field]

[0001] The present invention relates to an abnormality detection device and an abnormality detection method. [Background technology]

[0002] As an example of an anomaly detection device and anomaly detection method that can improve the accuracy of detecting rolling anomalies in plate compared to conventional methods, Patent Document 1 describes a device and anomaly detection method that includes a camera that images the rolled material, which is the target for detecting rolling anomalies, while the rolled material is being rolled; a brightness reference value setting unit that extracts the rolled material from at least one image captured by the camera and determines a brightness reference value at which it is determined that there are no abnormalities in the rolled material from the brightness data of pixels within the range of the rolled material; and an image processing unit that uses one or more of the one or more captured images as comparison images, extracts the rolled material from each comparison image, and detects surface rolling anomalies based on the brightness difference between the brightness data within the range of the rolled material and the brightness reference value. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 7025608 Summary of the Invention [Problem to be solved by the invention]

[0004] In the above-mentioned Patent Document 1, the brightness values ​​of the colors of normal parts of the surface of a metal strip that has been rolled are obtained in advance as reference values ​​(reference R value, reference G value, reference B value), and while the metal strip is being rolled, the brightness values ​​(R value, G value, B value) of the pixels of an image of the surface of the metal strip taken by a camera are obtained, and the reference R value is subtracted from the R value, the reference G value from the G value, and the reference B value from the B value to obtain each brightness value difference, and rolling abnormalities are detected based on the relationship between the brightness value differences of two components of each of the brightness value differences.

[0005] In this case, the reflected light area where the illumination light is reflected on the plate surface makes the color of the plate surface appear to shine, and the R brightness value, G brightness value, and B brightness value are all close to the maximum brightness value (255), which is far from the reference brightness value for the color of normal parts of the metal strip plate surface. Therefore, the reflected light area is excluded from processing when determining whether there is a rolling abnormality.

[0006] Here, in the technology described in Patent Document 1, a reference luminance value of the color of a normal part of the surface of a metal strip plate is determined, and a part where the luminance value of the color of the surface of the metal strip plate has a certain difference compared to the reference luminance value is judged to be an abnormal part of rolling, i.e., a part where sheet squeezing has occurred. However, a part where illumination light is reflected on the sheet surface, i.e., a part called a reflected light area, is excluded from the judgment of an abnormal part of rolling, as it is not a part where sheet squeezing has occurred.

[0007] Of these, the areas where the illumination light was reflected on the plate surface appeared mainly along the plate width direction at the curved parts of the plate in the looper section, and so these areas were called reflected light areas, and these areas were excluded from the determination of rolling abnormalities.

[0008] However, after extensive research, the inventors have found that there are cases where areas on the plate surface where illumination light is reflected occur, such as outlying areas, outside of the reflected light areas.

[0009] In the technology described in Patent Document 1, the outlying areas are not judged as reflected light areas, and therefore the outlying areas are judged as rolling abnormalities caused by the occurrence of sheet narrowing, which has been observed to result in a false positive in the judgment of sheet narrowing, making it clear that there is room for improvement.

[0010] The present invention provides an abnormality detection device and an abnormality detection method that are capable of detecting rolling abnormalities in metal strips with higher accuracy than conventional methods. [Means for solving the problem]

[0011] The present invention includes a plurality of means for solving the above-mentioned problems, and one example thereof is an abnormality detection device for detecting rolling abnormalities on the surface of a metal band plate being rolled by a rolling mill, the abnormality detection device comprising: a camera for capturing an image of the metal band plate, which is the target of detection for the rolling abnormality, while the metal band plate is being rolled; a brightness reference value setting unit for extracting the metal band plate from at least one image captured by the camera, and determining a brightness reference value at which it is determined that there is no abnormality in the plate from brightness data of pixels within the range of the metal band plate; and an image processing unit for using one or more of the one or more captured images as comparison images, extracting the metal band plate from each comparison image, and detecting the rolling abnormality based on the brightness value difference between the brightness data of pixels within the range of the metal band plate extracted from the comparison image and the brightness reference value, the image processing unit setting a detection area of ​​the metal band plate, which is the target of detection for the abnormality, from the latest image, and determining a plurality of reflected light areas on the surface of the metal band plate, which are caused by illumination light illuminating the metal band plate. a first region of reflected light extending in the width direction of the metal band plate, and a second region of reflected light separated from the first region; a new detection area is set, and the area obtained by removing the plurality of reflected light areas from the detection area is set as a new detection area, and an abnormality area is set based on the relationship between the brightness value difference between the brightness data of the pixels in the new detection area and the brightness reference value, and the rolling abnormality is detected. [Effects of the Invention]

[0012] According to the present invention, rolling abnormalities in a metal strip can be detected with higher accuracy than in the past. Objects, configurations, and effects other than those described above will become apparent from the following description of the embodiments. [Brief explanation of the drawings]

[0013] [Figure 1] 1 is a schematic diagram showing the configuration of a rolling facility equipped with an abnormality detection device according to an embodiment of the present invention. [Figure 2] 10 is a diagram showing an overview of a state in which an outlying reflected light region occurs in addition to a band-shaped reflected light region on the plate surface of a metal band plate; [Figure 3] FIG. 3 is a diagram showing an overview of a state in which the outlying reflected light areas shown in FIG. 2 cannot be excluded, and the reflection of illumination light is mistakenly detected as a narrowing of the board. [Figure 4]5A and 5B are diagrams illustrating the flow of a method for calculating an abnormal area based on a reflected light area in the abnormality detection device of the embodiment. [Figure 5] 5A and 5B are diagrams illustrating the flow of a method for calculating an abnormal area based on a reflected light area in the abnormality detection device of the embodiment. [Figure 6] FIG. 10 is a diagram showing an example of chromaticity of a normal surface of a metal strip. [Figure 7] 5A and 5B are diagrams illustrating the flow of a method for calculating an abnormal area based on a reflected light area in the abnormality detection device of the embodiment. [Figure 8] 4 is a flowchart of an abnormality determination process in the abnormality detection device of the embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0014] An embodiment of the anomaly detection device and anomaly detection method of the present invention will be described with reference to Figures 1 to 8. In the drawings used in this specification, identical or corresponding components are denoted by the same or similar reference numerals, and repeated explanations of these components may be omitted.

[0015] The metal strip to be rolled in the present invention is generally a strip of a metal material that can be rolled, and the type is not particularly limited, and in addition to steel plate, it can be a non-ferrous material such as aluminum or copper.

[0016] First, the overall configuration of the rolling facility including the abnormality detection device and the configuration of the abnormality detection device will be described with reference to Fig. 1. Fig. 1 is a schematic diagram showing the configuration of the abnormality detection device of this embodiment and the rolling facility equipped with it.

[0017] The rolling facility 100 shown in FIG. 1 is a finishing rolling facility for rolling a metal strip 1, and is equipped with an F1 stand 10, an F2 stand 20, an F3 stand 30, an F4 stand 40, an F5 stand 50, an F6 stand 60, an F7 stand 70, cameras 81 and 82, a looper 65 for tension control, an image processing computer 90, a database 93, a display device 95, etc.

[0018] Of these, the cameras 81 and 82 and the image processing computer 90 constitute an abnormality detection device that detects rolling abnormalities on the surface of the metal strip 1 being rolled by the rolling mill.

[0019] The rolling equipment 100 is not limited to a configuration in which seven rolling stands are provided as shown in FIG. 1, but may have at least one stand.

[0020] Each of the F1 stand 10, F2 stand 20, F3 stand 30, F4 stand 40, F5 stand 50, F6 stand 60, and F7 stand 70 is a rolling mill equipped with an upper work roll, a lower work roll, an upper backup roll that supports the upper work roll and the lower work roll by contacting them, a lower backup roll, reduction cylinders 11, 21, 31, 41, 51, 61, 71 provided above the upper backup roll, and load detectors 12, 22, 32, 42, 52, 62, 72. A six-stage configuration may be achieved by providing upper and lower intermediate rolls between each of the upper and lower work rolls and each of the upper and lower backup rolls. The roll configuration of the rolling mill is not limited to the above, and it is sufficient to have at least upper and lower work rolls.

[0021] The looper 65 is a roll for controlling the line tension. This roll is installed so that its rotation axis extends in the width direction of the metal strip 1, and is installed between the F1 stand 10 and the F2 stand 20, between the F2 stand 20 and the F3 stand 30, between the F3 stand 30 and the F4 stand 40, between the F4 stand 40 and the F5 stand 50, between the F5 stand 50 and the F6 stand 60, and between the F6 stand 60 and the F7 stand 70 so that the line tension can be changed by raising or lowering the metal strip 1 in the vertical direction. The looper 65 may also have the function of a strip shape meter that detects the tension distribution in the strip width direction.

[0022] The camera 81 is provided at a position where it can capture images including the metal strip 1 on the exit side of the F4 stand 40 and the entrance side of the F5 stand 50, and captures images including the metal strip 1 from directly above the metal strip 1 or from diagonally above at intervals shorter than 0.1 seconds, preferably in video format. The image data captured by the camera 81 is transmitted to the image processing computer 90 via a communication line 85.

[0023] The camera 82 is provided on the exit side of the F7 stand 70 at a position where it can capture images including the metal strip 1, and like the camera 81, it captures images including the metal strip 1, which is the target for detecting rolling abnormalities, from directly above the metal strip 1 or from diagonally above at intervals shorter than 0.1 seconds, for example. The data of the images captured by the camera 82 is also transmitted to the image processing computer 90 via the communication line 85.

[0024] An imaging step is performed by these cameras 81 and 82 to image the metal strip 1, which is the target for detecting rolling abnormalities, while the metal strip 1 is being rolled.

[0025] In this embodiment, these cameras 81 and 82 capture images of the metal strip 1 at different times during rolling of one roll that constitutes the metal strip 1, which is the target of abnormality detection.

[0026] In the following description, cameras are installed at two locations: at the exit side of F4 stand 40, the entry side of F5 stand 50, and the exit side of F7 stand 70. However, cameras only need to be installed at one location, and there can be one or more cameras, and they can also be installed between all stands or at the entry and exit sides of the rolling equipment 100.

[0027] The lighting device 67, not shown, illuminates the metal strip 1, particularly the metal strip 1 in the range imaged by the cameras 81 and 82, and can be a general lighting device appropriately placed on the ceiling of the rolling factory where the rolling equipment 100 is installed, or it can be a dedicated lighting device.

[0028] The image processing computer 90 is a device made up of a computer and the like that controls the operation of each device in the rolling facility 100, and has a brightness reference value setting unit 91, an image processing unit 92, and the like.

[0029] The brightness reference value setting unit 91 is a part that extracts the metal band plate 1 from at least one image captured by the cameras 81 and 82, and determines a brightness reference value at which it is determined that there is no abnormality in the plate from the brightness data of pixels within the range of the metal band plate 1, and is the main body that executes the brightness reference value setting process. When a moving image is captured by the cameras 81 and 82, the brightness reference value setting unit 91 extracts an image from the moving image.

[0030] The image processing unit 92 is a part that performs the image processing step, selecting one or more of the captured images as a comparison image, extracting the metal strip 1 from each comparison image, and detecting rolling abnormalities based on the difference in brightness between the brightness data of pixels within the range of the metal strip 1 extracted from the comparison image and a brightness reference value. The image processing unit 92 includes a detection area setting unit 92A, a reflected light area setting unit 92B, an abnormal area setting unit 92C, etc. It is desirable that the comparison image processed by this image processing unit 92 be basically the most recent image.

[0031] The detection area setting section 92A is a section that sets a detection area 120 (see FIG. 4) of the metal band plate 1, which is a target for detecting abnormalities, from the latest image.

[0032] The reflected light region setting section 92B is a section for setting a reflected light region 101 and an isolated reflected light region 103 (see FIG. 4) reflected on the surface of the metal band 1 due to the illumination light illuminating the metal band 1.

[0033] The abnormality area setting unit 92C is a part that sets the area obtained by removing the reflected light area 101 and the outlying reflected light area 103 from the detection area 120 as a new detection area, and sets abnormal areas 111 and 113 based on the relationship between the brightness value difference between the brightness data of the pixels in the new detection area and the brightness reference value, thereby detecting rolling abnormalities.

[0034] The details of the processes in the brightness reference value setting unit 91, image processing unit 92, detection area setting unit 92A, reflected light area setting unit 92B, and abnormal area setting unit 92C will be described later.

[0035] The database 93 is a storage device that stores information on threshold boundaries that separate normal and abnormal values, which have been previously calculated from a distribution of the brightness values ​​of one component plotted against the brightness value difference of the other component on a two-dimensional graph consisting of two components, among the R (red), G (green), and B (blue) brightness values ​​of pixels within the range of the metal strip 1 in an image that captures the metal strip 1. It is preferably configured as an SSD, HDD, or the like.

[0036] The display device 95 is a display device such as a display or an audio device such as an alarm, and is a device for informing an operator of the occurrence of a strip shape abnormality and the work to be done to deal with it when the image processing computer 90 determines that the strip shape is abnormal. For this reason, a display is often used as the display device 95.

[0037] During operation, the operator checks for the presence or absence of rolling abnormalities such as the rolled strip squeezing phenomenon by visually checking the display screen of the display device 95, each stand itself, and the gaps between each stand. For example, if the operator checks the display device 95 to see that the rolled strip squeezing phenomenon has occurred, and if the display shows operations that should be performed, such as bending correction, leveling correction, roll speed correction, and opening the roll gap of the downstream stand, the operator can manually perform the operations in accordance with the instructions, thereby preventing the abnormal rolling state from progressing to a further worsening state.

[0038] Incidentally, the occurrence of the abnormality of the phenomenon of the rolled sheet squeezing is displayed on the display screen of the display device 95 for the operator, and a signal to correct or stop the rolling to avoid the rolling abnormality is sent to the rolling mill control device, so that the rolling abnormality avoidance operation is performed by automatic control, or it can be configured so that various corrective operations or stop operation are performed by automatic control without displaying on the display device 95.

[0039] The image processing computer 90, database 93, and rolling mill control device described above can be configured as a computer having a display device 95 such as a liquid crystal display (described later), an input device, a storage device, a CPU, a memory, etc., and these may be configured as a single computer, or each may be configured as a separate computer, and are not particularly limited.

[0040] The image processing computer 90 and the rolling mill control device control the operation of each device based on various programs recorded in a storage device. The control processes for the operations executed by the image processing computer 90 and the rolling mill control device may be integrated into a single program, or may be separated into multiple programs, or a combination of these. Furthermore, some or all of the programs may be realized by dedicated hardware or may be modularized.

[0041] Next, specific examples of the anomaly detection process for determining abnormalities in the shape of the rolled metal strip 1 according to the present invention will be described with reference to Figures 2 to 7. Figure 2 is a diagram showing an overview of a state in which outlying reflected light areas occur in addition to band-shaped reflected light areas on the surface of the metal strip, Figure 3 is a diagram showing an overview of a state in which the outlying reflected light areas shown in Figure 2 cannot be excluded and the reflection of the illumination light is erroneously detected as the occurrence of strip narrowing, Figures 4, 5 and 7 are diagrams explaining the flow of a method for calculating an abnormal area from a reflected light area in the anomaly detection device of the embodiment, and Figure 6 is a diagram showing an example of the chromaticity of a normal surface of a metal strip.

[0042] In the technology described in the above-mentioned Patent Document 1, the reflection points of the illumination light that appear on the surface of the metal strip 1 along the strip width direction, mainly at the curved parts of the strip at the looper 65, are called reflected light areas 2, and other areas on the strip surface where the color is far from the normal color are determined to be abnormal parts of the rolling, i.e., areas where strip narrowing has occurred.

[0043] However, after extensive research, the inventors discovered that there are cases where illumination light is reflected in areas other than the area identified as the reflected light area 2, such as isolated reflected light areas 3, as shown in Figure 2.

[0044] Although this outlying reflected light region 3 is not a rolling abnormality, it was not excluded as a reflected light region as shown in Figure 3, but was instead determined to be an abnormality when determining abnormalities based on brightness values, which revealed that there is room for improvement in the reliability of abnormality determination. Based on this finding, the inventors came up with the idea of ​​excluding such outlying reflected light region 3, in addition to the band-like reflected light region 2, as a reflection point of illumination light from the determination of rolling abnormalities, and completed the present invention.

[0045] In the present invention, too, the process of determining whether or not there is an abnormal part in the rolling after removing the reflected light region 2 and the outlying reflected light region 3, i.e., a location where strip narrowing has occurred, is the same as in Patent Document 1 described above, in which the brightness values ​​of the color of normal parts of the surface of the metal strip 1 are obtained as reference values ​​(reference R value, reference G value, reference B value), brightness values ​​(R value, G value, B value) of pixels in images of the surface of the metal strip 1 taken by cameras 81 and 82 during the rolling of the metal strip 1 are obtained, the reference R value is subtracted from the R value, the reference G value is subtracted from the G value, and the reference B value is subtracted from the B value to obtain each brightness value difference, and the rolling abnormal part is identified based on the relationship between the brightness value differences of two components of each brightness value difference.

[0046] The details of the process will be explained below.

[0047] First, cameras 81 and 82 capture images including the metal strip 1. Preferably, the images of the metal strip 1 are captured continuously in a moving image format. The captured images are output to an image processing computer 90 via a communication line 85.

[0048] In the image processing computer 90, the brightness reference value setting unit 91 uses a binarization process to distinguish between the background and the area where the metal strip 1 exists from the image of the metal strip 1 during rolling, and extracts a range 1A of the metal strip 1 from the area with a brightness above a certain threshold.

[0049] Thereafter, the extracted range 1A is subjected to abnormality detection processing in the image processing unit 92, which includes a detection area setting unit 92A, a reflected light area setting unit 92B, an abnormal area setting unit 92C, and the like.

[0050] For example, consider the case where an image is captured that includes a reflected light region 101 having a band-like reflected light and a protrusion 101a partially extending in the rolling direction, an isolated reflected light region 103, and abnormal regions 111 and 113 where rolling abnormalities have occurred, as shown in Figure 4.

[0051] First, the brightness reference value setting unit 91 extracts a range 1A of the metal strip 1, and then the image processing unit 92 extracts a reflected light region 101 and an isolated reflected light region 103 within this range 1A using a first predetermined reference value and a second predetermined reference value, as shown in Figure 5.

[0052] In excluding the reflected light regions in this embodiment, instead of excluding only specific regions known as reflected light regions as in Patent Document 1, the areas where the color of the plate surface of the metal strip plate 1 appears to shine and where the R brightness value, G brightness value, and B brightness value are all close to the maximum brightness value (255) are regarded as reflecting areas of the illumination light and are excluded from the targets for determining rolling abnormalities.

[0053] The color of a normal surface of a hot-rolled steel sheet is shown in Figure 6. As shown in Figure 6, the color of a normal surface of a steel sheet that constitutes a metal strip 1 during hot rolling has the highest R value (red), while the G value (green) and B value (blue) are relatively low.

[0054] Therefore, in this case, the brightness values ​​of the G value (green) and B value (blue), which indicate low chromaticity, are examined, and if the brightness value of the G value is 250 or higher, for example, it can be determined that the area is reflecting illumination light.

[0055] In this way, when the metal strip plate 1 is a steel plate, the image processing unit 92 divides the brightness data of pixels within the range of the metal strip plate 1 extracted from the comparison image into R values, G values, and B values, and can set as multiple reflected light areas the areas where the brightness value of at least one component of the G value or B value is equal to or greater than a first predetermined reference value.

[0056] The first predetermined reference value is set so that if the brightness value other than the high brightness R value is 250 or more, it can be reliably determined to be a reflection point of the illumination light, but "250" is only an example, and it is desirable to change and set the first predetermined reference value as appropriate after investigating the operating environment of the rolling equipment at the preparation stage for operation of the rolling equipment.

[0057] Furthermore, if the metal strip 1 is a hot-rolled material other than a steel plate, the chromaticity distribution will not be the same as that shown in Figure 6. In such cases, it is preferable to determine the location as a reflection of the illumination light under the condition that the brightness values ​​of R (red), G (green), and B (blue) are all, for example, 250 or higher.

[0058] In this way, when the metal strip 1 is made of a material other than steel, such as aluminum or copper, the image processing unit 92 separates the brightness data of pixels within the range of the metal strip 1 extracted from the comparison image into R, G, and B values, and can set, as multiple reflected light regions, portions where the brightness values ​​of all the R, G, and B components are equal to or greater than the second predetermined reference value. Note that, like the first predetermined reference value, this second predetermined reference value, "250," is merely an example, and it is desirable to change and set it appropriately after investigating the operating environment of the rolling equipment and the type of metal strip 1 to be rolled during the preparation stage for operation of the rolling equipment.

[0059] In this way, the image processing unit 92 can set multiple reflected light regions, including a reflected light region 101, which is reflected light extending in the width direction of the metal band plate 1, and an isolated reflected light region 103, which is reflected light separated from the reflected light region 101.

[0060] Here, the inventors' investigations revealed that the outlying reflected light region 103 was very often located upstream of the reflected light region 101 in the rolling direction. Therefore, by setting the outlying reflected light region 103 as being located upstream of the reflected light region 101 in the rolling direction, it becomes possible to take measures such as narrowing the search region, thereby speeding up processing.

[0061] As a result, only true abnormal regions (regions that are relatively long in the conveyance direction) are extracted, as shown in Fig. 7. Thereafter, the detected true abnormal regions are set as abnormal regions 111 and 113 by the abnormal region setting unit 92C, and the process proceeds to final detection processing for determining whether or not a rolling abnormality has been detected.

[0062] The process of setting the detection area 120 of the metal strip 1 to be detected for abnormalities and the process of setting the multiple reflected light areas 101 and isolated reflected light areas 103 reflected on the surface of the metal strip 1 due to the illumination light illuminating the metal strip 1 can be performed in any order, and they can be performed in parallel or one can be performed first, and there is no particular limitation.

[0063] Thereafter, the image processing unit 92 determines that a rolling abnormality has occurred in the metal strip 1 if the total area value of the abnormal area set by removing the reflected light area from the lighting device 67 using the method of Patent Document 1 or the like or the maximum length in the rolling direction of the abnormal area is greater than each specific threshold value.

[0064] The extracted range 1A may include an area slightly inside or outside the actual plate area, and does not necessarily have to coincide with the metal strip plate 1.

[0065] In the image processing computer 90, the brightness reference value setting unit 91 determines the brightness distribution of pixels inside the range 1A of the metal strip plate 1 for one selected image and calculates a brightness reference value at which the plate is determined to be normal. Here, the brightness reference value setting unit 91 separates the brightness data into three components: R value, G value, and B value, and determines three brightness reference values: a reference R value (R0 value), a reference G value (G0 value), and a reference B value (B0 value).

[0066] It should be noted that two of the brightness reference values ​​(any combination of the R0 value and G0 value, the R0 value and B0 value, or the B0 value and G0 value) may be obtained.

[0067] Furthermore, the brightness reference value setting unit 91 can be configured to process brightness data in grayscale with 8 bits (256 gradations) or 16 bits (65536 gradations), etc. In the case of grayscale, the subsequent processing is basically the same, and details will be omitted.

[0068] The three brightness reference values, the reference R value (R0 value), the reference G value (G0 value), and the reference B value (B0 value), can be calculated by selecting them from one image or by selecting them from multiple images. Details of these methods can be found in, for example, the method described in Patent Document 1, and will not be discussed further here.

[0069] Thereafter, the image processing unit 92 detects rolling abnormalities based on the relationship between the brightness value differences of two components among the respective brightness value differences. There are two methods for determining abnormalities: (i) determination based on the relationship between the brightness value differences of two components (using a specific threshold value), and (ii) determination based on the relationship between the brightness value differences of two components (using an arbitrary threshold boundary). The details of these methods can also be determined using the method described in Patent Document 1, and so further details will be omitted here.

[0070] Next, an abnormality detection method for detecting abnormalities on the surface of the metal strip 1 rolled by the rolling mill according to this embodiment will be described with reference to Fig. 8. Fig. 8 is a flowchart of an abnormality determination process in the abnormality detection device according to this embodiment.

[0071] First, as shown in Fig. 8, images are acquired by the cameras 81 and 82 (step S101). This step S101 corresponds to an imaging step, and preferably, a moving image is captured.

[0072] Next, the brightness reference value setting unit 91 of the image processing computer 90 determines whether or not the metal band plate 1 is present in the image captured in step S101 (step S102). If it is determined that the metal band plate 1 is present, the process proceeds to step S103, whereas if it is determined that the metal band plate 1 is not present, the process proceeds to step S111.

[0073] Thereafter, the brightness reference value setting unit 91 sets a plate detection area (step S103), and then selects a brightness reference value for each target coil (step S104). Steps S103 and S104 correspond to a brightness reference value setting step.

[0074] Next, the image processing unit 92 of the image processing computer 90 calculates the difference in brightness between the brightness reference value selection image and a comparison image captured after the brightness reference value selection image (step S105), and determines whether the image is an abnormality candidate (step S106). The image processing unit 92 also performs processing to remove reflected light areas originating from the lighting device 67 (step S107). Note that, as described above, the order of steps S106 and S107 is not critical, and either may precede the other, or the two may be processed in parallel.

[0075] Thereafter, the image processing unit 92 executes a final abnormality determination process (step S108), and determines whether or not a rolling abnormality such as drawing has occurred (step S109). If it is determined that an abnormality has occurred, the process proceeds to step S110, and if it is determined that no abnormality has occurred, the process proceeds to step S111. These steps S105 to S109 correspond to the image processing step.

[0076] If it is determined in step S110 that an abnormality has occurred, the image processing unit 92 records the abnormality occurrence flag as 1 (step S110), and if it is determined in step S102 that the metal strip plate 1 is not present or if it is determined in step S110 that no abnormality has occurred, the image processing unit 92 records the abnormality occurrence flag as 0 (step S111), and starts processing again at the next timing.

[0077] When the image processing computer 90 records the abnormality occurrence flag 1, it displays that fact on the display device 95. Alternatively, it is possible to automatically execute intervention processing for each of the stands 10, 20, 30, 40, 50, 60, and 70.

[0078] The above-described processing of steps S101 to S111 is executed during the rolling of one rolled coil, and the processing ends when the rolling of one rolled coil is completed.

[0079] Next, the effects of this embodiment will be described.

[0080] The abnormality detection device for detecting rolling abnormalities on the surface of a metal strip 1 rolled by the rolling mill of this embodiment described above comprises cameras 81, 82 (imaging step) for capturing images of the metal strip 1 during rolling of the metal strip 1 that is the target for detecting rolling abnormalities, a brightness reference value setting unit 91 (brightness reference value setting step) for extracting the metal strip 1 from at least one image captured by the cameras 81, 82 and determining a brightness reference value at which it is determined that there is no abnormality in the strip from brightness data of pixels within the range of the metal strip 1, and a brightness reference value setting unit 92 (brightness reference value setting step) for selecting one or more of the one or more captured images as comparison images, extracting the metal strip 1 from each comparison image, and determining a brightness reference value at which it is determined that there is no abnormality in the strip from brightness data of pixels within the range of the metal strip 1 extracted from the comparison image. and an image processing unit 92 (image processing step) that detects rolling abnormalities based on the brightness value difference between the brightness data and the brightness reference value. The image processing unit 92 sets, from the latest image, a detection area 120 of the metal strip 1 that is to be detected for abnormalities, and also sets a reflected light area 101 and an outlying reflected light area 103 that are reflected on the surface of the metal strip 1 and are caused by the illumination light that illuminates the metal strip 1. The image processing unit 92 removes the reflected light area 101 and the outlying reflected light area 103 from the detection area 120 and sets the area as a new detection area, and sets abnormal areas 111, 113 based on the relationship between the brightness value difference between the brightness data of the pixels in the new detection area and the brightness reference value, thereby detecting rolling abnormalities.

[0081] Therefore, it is possible to exclude the outlying reflected light area 103, which may have been judged as a rolling abnormality area even though there was no rolling abnormality, from the abnormality judgment process, thereby achieving an improvement in the accuracy of the abnormality detection process for the metal strip plate 1 compared to the conventional method.

[0082] Furthermore, the image processing unit 92 sets, as a plurality of reflected light regions, a reflected light region 101 which is reflected light extending in the width direction of the metal band plate 1 and an isolated reflected light region 103 which is reflected light separated from the reflected light region 101. abnormalityThe detection accuracy can be further improved.

[0083] Furthermore, the image processing unit 92 can speed up the detection process of the isolated reflected light region 103 by setting the isolated reflected light region 103 as being located upstream of the reflected light region 101 in the rolling direction.

[0084] In addition, the image processing unit 92 divides the brightness data of pixels within the range of the metal strip 1 extracted from the comparison image into R values, G values, and B values, and sets areas where the brightness value of at least one of the G value and B value components is equal to or greater than a first predetermined reference value as multiple reflected light areas, thereby making it possible to extract reflected light areas with very high accuracy when the metal strip 1 is a steel plate.

[0085] Furthermore, the image processing unit 92 divides the brightness data of pixels within the range of the metal strip 1 extracted from the comparison image into R values, G values, and B values, and sets areas where the brightness values ​​of all R, G, and B components are equal to or greater than a second predetermined reference value as multiple reflected light areas, thereby making it possible to extract reflected light areas with very high accuracy even when the metal strip 1 is not a steel plate.

[0086] <Other> The present invention is not limited to the above-described embodiment, and various modifications and applications are possible. The above-described embodiment has been described in detail to clearly explain the present invention, and the present invention is not necessarily limited to having all of the described configurations. [Explanation of symbols]

[0087] 1...Metal strip 1A...range 2,101...Reflected light area (first area) 3,103... Outlying reflected light area (second area) 10...F1 Stand 11, 21, 31, 41, 51, 61, 71...Pressure cylinder 12, 22, 32, 42, 52, 62, 72...Load detector 20...F2 Stand 30...F3 Stand 40...F4 Stand 50…F5 Stand 60...F6 Stand 65...Looper 67...Lighting equipment 70…F7 Stand 81, 82...Camera 85...Communication line 90...Image processing computer 91...Luminance reference value setting unit 92...Image processing unit 92A…Detection area setting section 92B...Reflected light area setting section 92C…Abnormal area setting section 93...Database 95...Display device 100...Rolling equipment 101a...Protruding part 111,113…Abnormal area 120…detection area

Claims

1. An abnormality detection device for detecting rolling abnormalities on the surface of a metal strip being rolled by a rolling mill, a camera that captures an image of the metal strip during rolling of the metal strip that is the target of the rolling abnormality detection; a brightness reference value setting unit that extracts the metal band plate from at least one image captured by the camera and determines a brightness reference value at which the plate is judged to be normal based on brightness data of pixels within the range of the metal band plate; and an image processing unit that uses one or more of the one or more captured images as a comparison image, extracts the metal strip from each comparison image, and detects the rolling abnormality based on a brightness value difference between brightness data of pixels within the range of the metal strip extracted from the comparison image and the brightness reference value, The image processing unit A detection area of ​​the metal band plate that is to be detected for abnormalities is set from the latest image, and a plurality of reflected light areas that are reflected on the surface of the metal band plate due to illumination light that illuminates the metal band plate are set, including a first area that is reflected light extending in the width direction of the metal band plate and a second area that is reflected light separated from the first area; A region obtained by removing the plurality of reflected light regions from the detection region is newly set as a new detection region, and an abnormality region is set based on the relationship between the brightness data of pixels in the new detection region and the brightness value difference of the brightness reference value, thereby detecting the rolling abnormality. Anomaly detection device.

2. 2. The abnormality detection device according to claim 1, The image processing unit sets the second region as being located upstream of the first region in the rolling direction. Anomaly detection device.

3. 3. The abnormality detection device according to claim 1, The image processing unit divides the luminance data of pixels within the range of the metal band plate extracted from the comparison image into R values, G values, and B values, and sets portions where the luminance value of at least one of the G values ​​and the B values ​​is equal to or greater than a first predetermined reference value as the plurality of reflected light regions. Anomaly detection device.

4. 3. The abnormality detection device according to claim 1, The image processing unit divides the luminance data of pixels within the range of the metal band plate extracted from the comparison image into R values, G values, and B values, and sets portions where the luminance values ​​of all components of the R values, G values, and B values ​​are equal to or greater than a second predetermined reference value as the plurality of reflected light regions. Anomaly detection device.

5. An abnormality detection method for detecting rolling abnormalities on the surface of a metal strip being rolled by a rolling mill, comprising: an imaging step of imaging the metal strip with a camera while the metal strip is being rolled; a brightness reference value setting step of extracting the metal band plate from at least one image captured in the imaging step and determining a brightness reference value at which the plate is judged to be normal based on brightness data of pixels within the range of the metal band plate; an image processing step of using one or more of the one or more captured images as a comparison image, extracting the metal strip from each comparison image, and detecting the rolling abnormality based on a brightness value difference between brightness data of pixels within the range of the metal strip extracted from the comparison image and the brightness reference value, In the image processing step, A detection area of ​​the metal band plate that is to be detected for abnormalities is set from the latest image, and a plurality of reflected light areas that are reflected on the surface of the metal band plate due to illumination light that illuminates the metal band plate are set, including a first area that is reflected light extending in the width direction of the metal band plate and a second area that is reflected light separated from the first area; A region obtained by removing the plurality of reflected light regions from the detection region is newly set as a new detection region, and the new detection region is set as an abnormal region based on the relationship between the brightness data of pixels in the new detection region and the brightness value difference of the brightness reference value, thereby detecting the rolling abnormality. Anomaly detection methods.

Images