TWI566848B - Error detection method for detecting a steel strip tail - Google Patents

Description

Method for detecting abnormality of rolling end of steel strip

The invention relates to a method for detecting an abnormality of rolling end of a steel strip.

In the current steel strip production process, after the steel embryo is heated by the heating furnace, the rough rolling and finishing rolling steps are carried out to obtain a steel strip having a desired thickness. For the convenience of transportation, the steel strip will be rolled into a coil by a coiler to facilitate transportation. In the rolling process of the production line, the steel strip may have a side walk, which may cause abnormal defects such as the folded edge of the steel strip. When the steel strip enters the rolling implement, the tail end of the steel strip will impact the roll of the rolling implement, and the surface of the roll will be uneven. This uneven defect causes the roll to transfer the defect on the surface of the roll to the steel strip when the next steel strip is rolled, which will cause the strip of the strip which is subsequently rolled by the roll to form a defect and be rejected. This abnormality in the transfer of the surface defects of the rolls due to the abnormal rolling at the end of the steel strip is called Tail Pinch.

In order to solve the above problems, a method for detecting the abnormality of the end rolling of the steel strip is needed to reduce the occurrence of the rolling transfer and the deviation of the steel strip by monitoring the condition of the end of the steel strip.

One aspect of the present invention is to provide an abnormality of rolling end of a steel strip The detection method is to monitor the condition of the end rolling of the steel strip and thereby analyze whether an abnormality may occur.

According to an embodiment of the present invention, in the method for detecting the abnormality of the end rolling of the steel strip, the steel strip image is first obtained, wherein the steel strip images are time continuous images. Next, the steel strip feature calculation step is performed for each steel strip image. In the steel strip feature calculation step, the target image is first converted into a gray scale image, wherein the target image is one of the steel strip images. Secondly, the grayscale image is binarized to obtain a binarized image. Then, an object detecting step is performed to find the largest object in the binarized image. Next, a bounding rectangle is defined according to the large object in the binarized reference image. Then, an area calculation step is performed to calculate the area of the largest object. Next, a coordinate calculation step is performed to calculate the center point coordinates of the largest object. After the steel strip feature calculation step, the steel strip tail end image determination step is then performed. The image judgment step of the tail end of the steel strip is judged according to the center point coordinate of the largest object of the first steel strip image, the center point coordinate of the largest object of the second steel strip image, and the area of the largest object of the first steel strip image. Whether the second steel strip image is an outbound image of the steel strip tail end, wherein the first steel strip image and the second steel strip image are consecutive ones of the foregoing steel strip images. When the second steel strip image is judged as the outbound image of the steel strip tail end, an abnormality detecting step is performed to determine whether the end rolling of the steel strip is abnormal according to the second steel strip image. For example, it is detected whether there is an offset at the end of the steel strip or whether a Tail Pinch may occur.

It can be seen from the above description that the method for detecting the abnormality of the end rolling of the steel strip in the embodiment of the present invention can judge the steel by the characteristic value of the outbound image of the steel strip tail end. Whether there is an abnormal condition at the end of the belt. In this way, the occurrence of roll transfer and strip offset can be greatly reduced.

200‧‧‧ Detection method for abnormal rolling end of steel strip

210‧‧‧Image capture steps

220‧‧‧ Steel strip feature calculation steps

221‧‧‧Color separation steps

222‧‧‧ Binarization steps

223‧‧‧Smoothing steps

224‧‧‧Object detection steps

225‧‧‧ Calculate eigenvalues

240‧‧‧Steel belt centerline calculation steps

250‧‧‧ Steel strip width calculation steps

260‧‧‧ Steel strip tail end image judgment step

261‧‧‧Horizontal coordinate difference calculation steps

262-266‧‧‧ Judgment steps

270‧‧‧Anomaly detection steps

272‧‧‧ Steel strip offset evaluation steps

272a‧‧‧Image selection steps

272b‧‧‧End offset calculation steps

272c‧‧‧ judgment steps

272d‧‧‧ Steel strip upper edge amplitude calculation steps

272e‧‧‧ Steel belt lower edge amplitude calculation steps

272f‧‧‧Upper edge amplitude judgment step

272g‧‧‧ Lower edge amplitude judgment step

274‧‧‧Wedge ratio calculation steps

274a‧‧‧Edge detection steps

274b‧‧‧ contour point definition steps

274c‧‧‧ Calculation steps

274d‧‧‧ judgment steps

600‧‧‧ steel strip tail profile

C1-C4‧‧‧ image capture device

F1-F4‧‧‧ rolling mill station

M‧‧‧ steel strip

OB‧‧‧ objects

ROI‧‧‧region of interest

S‧‧‧Maximum object

US‧‧‧Upper Edge Point Group

LS‧‧‧ lower edge point group

W‧‧‧Width of the area of interest

H‧‧‧The length of the area of interest

w‧‧‧The width of the circumscribed rectangle

h‧‧‧The length of the circumscribed rectangle

The above and other objects, features, and advantages of the present invention will become more apparent and understood. A schematic view of the rolling of the steel strip of this embodiment at a rolling mill station.

FIG. 2 is a view showing a method for detecting a rolling end abnormality of a steel strip according to an embodiment of the present invention.

Fig. 3a is a diagram showing processed images after binarization and expansion processing according to an embodiment of the present invention.

Figure 3b is a diagram showing the largest object in a binarized image in accordance with an embodiment of the present invention.

Figure 3c illustrates a region of interest in a binarized image in accordance with an embodiment of the present invention.

Figure 3d is a diagram showing the upper edge point group and the lower edge point group of the steel strip according to an embodiment of the present invention.

Figure 3e is a diagram showing a region of interest of a steel strip image and a minimum circumscribed rectangle in accordance with an embodiment of the present invention.

FIG. 4 is a schematic flow chart showing the step of judging the image of the exit end of the steel strip according to the embodiment of the present invention.

Figure 5 is a flow chart showing the steps of the steel strip offset evaluation according to an embodiment of the present invention.

FIG. 6 is a flow chart showing a step of calculating a wedge ratio according to an embodiment of the present invention.

Figure 6a is a diagram showing a steel strip profile in a binarized image in accordance with an embodiment of the present invention.

Please refer to FIG. 1 , which is a schematic view showing the rolling of the steel strip according to the embodiment at the rolling mill station. The image capturing devices C1 - C4 are arranged between the rolling mill stations F1 - F4 to lie above the steel strip M. Take the image of the steel strip M. The image capturing devices C1-C4 are electrically connected to a computer device (not shown) to analyze the steel strip image between each roll station by using the computer device, and perform the end rolling of the steel strip according to the embodiment of the present invention. Abnormal detection method.

Referring to FIG. 2, a method 200 for detecting a rolling end abnormality of a steel strip according to an embodiment of the present invention is illustrated. Here, the image capturing device C2 will be described as an example, but the embodiment of the present invention is not limited thereto. In the method 200 for detecting the abnormality of the end rolling of the steel strip, the image capturing step 210 is first performed to obtain the steel strip image by using the image capturing device C2. Next, a steel strip feature calculation step 220 is performed to calculate image feature values of the steel strip image captured by the image capture device C2. In this embodiment, each image captured by the image capturing device and its corresponding feature value are stored in the computer device for use in other steps.

In the strip feature calculation step 220, a color separation step 221 is first performed to extract the red component of the target image (ie, the steel strip image being processed) to obtain a red target image (hereinafter referred to as a gray scale image). Because the steel strip is in Most of the rolling process is in a red hot state, so in this embodiment, the red component of the target image is taken out to help filter the objects other than the steel strip. However, embodiments of the invention are not limited to the color separation step 221. In other embodiments of the invention, grayscale images may also be obtained by other means. For example, instead of color separation, grayscale images are obtained directly from the brightness of each pixel of the steel strip image.

Next, a binarization step 222 is performed to binarize the grayscale image according to the preset grayscale threshold. Then, a smoothing step 223 is performed to make the edge of the object OB in the binarized image smoother by using the dilation processing rule in Morphology, as shown in Fig. 3a. However, embodiments of the invention are not limited to the smoothing step 223. In other embodiments of the invention, the smoothing step 223 may also be omitted if the user can accept a sawtooth condition in the image.

Then, an object detection step 224 is performed to find the largest object in the binarized image. In the present embodiment, the object detection step 224 uses the Blob extraction method to find the largest object S, as shown in Figure 3b. Next, a feature value calculation step 225 is performed to calculate the area of the largest object S, the minimum circumscribed rectangle of the largest object S, the central coordinate value of the largest object S, the contour of the largest object S, and the like.

In an embodiment of the invention, a Region Of Interest decision step may be selectively performed to determine the region of interest based on the smallest circumscribed rectangle of the largest object S. The region of interest ROI is used to reduce the computer resources required for subsequent image processing steps. However, in the present invention In other embodiments, if the user has sufficient computer resources, the determination step of the region of interest may also be omitted.

In the present embodiment, when the steel head end enters between the rolling mill stations F2 and F3, a Region Of Interest decision step is performed to define the region of interest for use in subsequent steps. As shown in Fig. 3c, the region of interest determination step of the present embodiment is determined based on the squareness of the largest object S. When the squareness of the largest object S meets the preset value, the ROI of the region of interest is defined as the range of the minimum circumscribed rectangle BR of the largest object S extending up and down and the left and right directions by a number of pixels, wherein the number of extended pixels can be according to the needs of the user. To decide.

Next, a strip centerline calculation step 240 is performed to calculate the coordinates of the centerline of the strip M. In this embodiment, the center point vertical coordinate (Y coordinate) of the largest object S in the region of interest ROI of the continuous _n binarized images is obtained, and is averaged and averaged to obtain the center of the steel strip. The Y coordinate of the line. In the present embodiment, n is 5, but the embodiment of the present invention is not limited thereto.

Then, a strip width calculation step 250 is performed to calculate the width of the strip M. In the present embodiment, the strip width calculation step 250 applies edge detection (for example, using Canny edge detector) to the largest object S in the region of interest ROI to set the point group US at the upper edge of the largest object S. (denoted as _{{u i, i = 1 to} n}) and the lower edge of the point cloud LS (denoted as _{{b j, j = 1 to} m}) removed, as shown in Fig. 3d. Then, let the sum of the absolute differences of the vertical coordinates of the upper edge point group US and the vertical coordinates of the lower edge point group average (ie, mean ( )) is the width of the steel strip.

Next, the steel strip tail end image determination step 260 is performed to determine whether the currently taken steel strip image is an outbound image of the steel strip tail end. In an embodiment of the invention, the image of the exit end of the steel strip is defined as the image of the end of the strip exiting the rolling mill station (e.g., mill station F2). The strip end exit image judging step 260 uses the ROI of the region of interest to determine whether the strip image satisfies the judging requirement of the exit end of the strip, thereby determining the outbound image of the strip end.

Please refer to FIG. 4 , which is a flow chart showing the step 260 of the exit end image of the steel strip. In the tail end image determination step 260 of the steel strip, the horizontal coordinate difference calculation step 261 is first performed to calculate the horizontal coordinate of the center point coordinate of the largest object of the previous image (hereinafter referred to as the first image) before capturing the image. The difference between the value and the horizontal coordinate value of the center point coordinate of the largest object of the currently captured image (hereinafter referred to as the second image). Next, a determining step 262 is performed to determine whether the difference is greater than a preset distance threshold and obtain a first determination result. In this embodiment, the preset distance threshold is 8.

Then, in a determining step 263, it is determined whether the horizontal coordinate value of the center point coordinate of the largest object of the second steel strip image is within the horizontal difference value to obtain the second determination result. In the present embodiment, the horizontal difference range is set to 0.2 W - 0.6 W, where W is the width of the ROI of the region of interest, as shown in Fig. 3e.

Next, in a determining step 264, it is determined whether the vertical coordinate value of the center point coordinate of the largest object of the second steel strip image is within the vertical difference value to obtain a third determination result. In this embodiment, the vertical difference range is set to 0.3H-0.8H, where H is the length of the region of interest ROI, as in the first Figure 3e shows.

Then, in a determining step 265, it is determined whether the object area of the largest object of the second steel strip image is greater than the area threshold to obtain a fourth determination result. In this embodiment, the area threshold is set to be 0.1 times the area of the largest object circumscribed rectangle, wherein the width of the circumscribed rectangle is w, the length is h, and the area is represented by w×h, as shown in FIG. 3e Show.

Next, in a determining step 266, it is determined whether the first to fourth determination results are all yes. If the first judgment result to the fourth judgment result are both yes, the second steel strip image (ie, the image currently captured) is judged as the outbound image of the steel strip tail end.

Returning to Fig. 1, after the exit image determination step 260 at the end of the steel strip, an abnormality detecting step 270 is performed to determine whether an abnormality may occur at the end of the steel strip by using the outbound image of the steel strip tail. In an embodiment of the present invention, the anomaly detection step 270 utilizes the exit image of the tail end of the steel strip to calculate the wedge ratio and evaluate the deviation of the steel strip for the user to take appropriate measures to avoid the steel strip. An abnormality occurs in the tail end of the tail pinch.

Please refer to FIG. 5, which is a schematic flow chart of the steel strip offset evaluation step 272 according to an embodiment of the present invention. The steel strip offset evaluation step 272 of the present embodiment determines whether the steel strip is offset according to the oscillation amplitude of the upper edge of the steel strip and the amplitude of the lower edge oscillation. In the strip offset evaluation step 272, an image selecting step 272a is first performed to select a plurality of strip images from the point of time corresponding to the exit image of the strip end. In the present embodiment, the image selection step 272a selects 40 consecutive strip images forward.

Then, a tail offset calculation step 272b is performed to calculate the steel strip trailing end offset value. In the present embodiment, the steel end offset calculation step 272b calculates the distance between the center coordinate of the largest object of the 41 steel strip images and the center line of the steel strip to obtain a plurality of steel strip tail end offset values. Next, a decision step 272c is performed to determine whether the maximum value of the tail end offset values of the steel strips is greater than a preset threshold. In the present embodiment, the preset threshold is 5, but the embodiment of the present invention is not limited thereto. When the maximum value of the tail end offset values of the steel strips is greater than the preset threshold value, the steel strip upper edge amplitude calculating step 272d and the steel strip lower edge amplitude calculating step 272e are respectively performed to determine whether the steel strip is up or down Offset.

The upper edge amplitude calculation step 272d of the steel strip is to calculate the upper edge oscillating amplitude of the steel strip image of the above 41 strips, wherein the upper edge of the steel strip is defined as the vertical coordinate difference between the upper edge of the steel strip and the center line of the steel strip, steel The upper edge of the band is defined as the upper edge of the largest object. Next, an upper edge amplitude determining step 272f is performed to determine whether the difference between the maximum value and the minimum value of the upper edge oscillating amplitude of the 41 strips of the steel strip image is greater than the upper edge oscillating amplitude threshold, wherein the upper edge oscillating amplitude threshold is 6 However, embodiments of the invention are not limited thereto.

The lower edge amplitude calculation step 272e of the steel strip is to calculate the amplitude of the lower edge oscillation of the steel strip image of the above 41 strips, wherein the lower amplitude of the steel strip is defined as the vertical coordinate difference between the lower edge of the steel strip and the center line of the steel strip, steel The lower edge is defined as the lower edge of the largest object. Next, the lower edge amplitude determining step 272g is performed to determine whether the difference between the maximum value and the minimum value of the lower edge oscillating amplitude of the 41 strips of the steel strip image is greater than the lower edge oscillating amplitude threshold, wherein the lower edge oscillating amplitude threshold is 6 However, embodiments of the invention are not limited thereto.

When the result of the determination of the upper edge amplitude determining step 272f is YES, the representative steel strip is excessively offset above the image, which may cause an abnormality. Similarly, when the result of the determination of the lower edge amplitude determining step 272g is YES, the representative steel strip is excessively offset below the image, which may also cause an abnormality. Therefore, when the result of the determination of the upper edge amplitude determining step 272f is YES or the result of the lower edge amplitude determining step 272g is YES, the computer device monitoring the rolling stand will issue a warning to inform the line personnel to take appropriate measures to avoid an abnormality.

Please refer to FIG. 6 , which is a flow chart showing a wedge ratio calculation step 274 according to an embodiment of the present invention. In the wedge ratio calculation step 274, an edge detection step 274a is first performed to define a strip end profile 600 in the exit image of the strip end using the edge detection algorithm, as shown in Figure 6a. Next, a contour point definition step 274b is performed to find the two contour points P1 and P2 at the steel strip trailing end contour 600. In the present embodiment, the wedge ratio calculation step 274 determines the contour points P1 and P2 according to the length h of the largest object circumscribed rectangle, wherein the coordinates of the contour point P1 are (0.15h, S1), and the contour point P2 The coordinates are (0.85h, S2). Then, a calculation step 274c is performed to calculate the wedge ratio using the contour points P1 and P2. Next, a determining step 274d is performed to determine whether the wedge ratio exceeds a preset wedge-shaped proportional threshold for the line personnel to evaluate the rolling condition of the steel strip tail end. For example, when the wedge ratio is too large, the wedge shape representing the rough rolling embryo is large, so the line personnel can adjust the leveling of the rough rolling and the finishing rolling to respond. For another example, when the rate of change of the wedge ratio of a finishing rolling mill station is high, it indicates that the rolling station has a high probability of occurrence of a roll pinch. If the steel coil continuously processed at the rolling mill station has a problem that the wedge ratio is too large, then Adjust the flatness of the finish to eliminate the change.

It can be seen from the above description that the method for detecting the abnormality of the end rolling of the steel strip in the embodiment of the present invention is to take the steel strip image of the rolling mill station and calculate the characteristics of each steel strip image to utilize each steel strip. The image feature values of the image are used to analyze the steel strip image to determine whether the steel strip is likely to be rolled and transferred. The method for detecting the abnormality of the end rolling of the steel strip in the embodiment of the present invention can issue an early warning to the online operator to improve the process setting to avoid the phenomenon of rolling transfer.

In addition, the method for detecting the abnormality of the end rolling of the steel strip of the embodiment includes a wedge ratio calculation step and a steel strip offset evaluation step. However, in other embodiments of the present invention, the method for detecting the abnormality of the end rolling of the steel strip may also include only the wedge ratio calculation step or the steel strip offset evaluation step according to actual needs. As such, the corresponding steps may be omitted or the order of execution may be changed. For example, the calculation step of the steel strip centerline can be performed in the strip offset evaluation step.

While the invention has been described above in terms of several embodiments, it is not intended to limit the scope of the invention, and the invention may be practiced in various embodiments without departing from the spirit and scope of the invention. The scope of protection of the present invention is defined by the scope of the appended claims.

200‧‧‧ Detection method for abnormal rolling end of steel strip

210‧‧‧Image capture steps

220‧‧‧ Steel strip feature calculation steps

221‧‧‧Color separation steps

222‧‧‧ Binarization steps

223‧‧‧Smoothing steps

224‧‧‧Object detection steps

225‧‧‧ Calculate eigenvalues

240‧‧‧Steel belt centerline calculation steps

250‧‧‧ Steel strip width calculation steps

260‧‧‧ Steel strip tail end image judgment step

270‧‧‧Anomaly detection steps

Claims (10)

A method for detecting a rolling end abnormality of a steel strip includes: obtaining a plurality of steel strip images, wherein the steel strip images are time continuous images; and performing a steel strip characteristic calculation step for each of the steel strip images, wherein The steel strip feature calculation step includes: converting a target image into a gray scale image, wherein the target image is one of the steel strip images; performing a binarization step on the gray scale image to obtain a binarization Image; performing an object detecting step to find a largest object in the binarized image; defining a bounding rectangle according to the largest object in the binarized reference image; performing an area calculation step Calculating the area of the largest object; performing a standard calculation step to calculate the coordinates of the center point of the largest object; performing a step of judging the outbound image of the steel strip to obtain the largest object according to a first steel strip image Determining whether the second steel strip image is the center point coordinate, the center point coordinate of the largest object of the second steel strip image, and the area of the largest object of the first steel strip image An outbound image with a tail end, wherein the first steel strip image and the second steel strip image are consecutive ones of the strip images; and when the second strip image is judged to be the end of the strip When performing images An abnormality detecting step is to determine whether the rolling end of the steel strip is abnormal according to the second steel strip image. The method for detecting the end rolling anomaly of the steel strip as described in claim 1 further includes performing a region of interest determining step to define a region of interest in the target image to enable the object detecting step Performing in the region of interest, wherein the region of interest determining step comprises: selecting a steel strip reference image, wherein the steel strip reference image is one of the steel strip images, and the steel strip enters the rolling mill station An image; and defining the region of interest based on the circumscribed rectangle of the largest object of the strip reference image. The method for detecting the abnormality of the end rolling of the steel strip according to Item 2 of the claim, wherein the step of judging the exit end of the steel strip comprises: calculating a center point coordinate of the largest object of the first steel strip image and a horizontal coordinate difference of a center point coordinate of the largest object of the second steel strip image; determining whether the horizontal coordinate difference value is greater than a preset distance threshold to obtain a first determination result; determining the second steel strip Whether the horizontal coordinate value of the center point coordinate of the largest object of the image is within a horizontal difference value to obtain a second determination result; determining whether the vertical coordinate value of the center point coordinate of the largest object of the second steel strip image is Located within a vertical difference range to obtain a third determination result; Determining whether an object area of the largest object of the second steel strip image is greater than an area threshold to obtain a fourth determination result; and when the first determination result, the second determination result, the third determination result, and the first When the results of all four determinations are yes, it is determined that the second steel strip image is an outbound image of the steel strip tail end. The method for detecting the abnormality of the end rolling of the steel strip according to the third item of claim 3, wherein the step of judging the exit end of the steel strip further comprises: selecting the external object according to the largest object in the image of the second steel strip Determining the horizontal difference range according to the width of the rectangle; determining the vertical difference range according to the length of the circumscribed rectangle of the largest object in the second steel strip image; and according to the largest object in the second steel strip image The area of the circumscribed rectangle determines the area threshold. The method for detecting the abnormality of the end rolling of the steel strip according to item 4 of the claim, wherein the horizontal difference ranges from 0.2 W to 0.6 W, and the vertical difference ranges from 0.3H to 0.8H, and the area threshold Is w×h×0.1, where W is the width of the region of interest, H is the length of the region of interest, and w is the width of the circumscribed rectangle of the largest object in the second strip image, h is the number The length of the circumscribed rectangle of the largest object in the image of the second steel strip. The method for detecting a rolling end abnormality of a steel strip according to Item 1 of the claim, wherein the abnormal detecting step comprises: An edge detection algorithm is used to define a steel strip tail end contour in the second steel strip image; two contour points are found on the steel strip tail end contour; and a wedge shape is calculated by using the coordinates of the two contour points The wedge ratio is used by online personnel to evaluate the probability of a tail pinch occurring at the end of the steel strip. The method for detecting a rolling end abnormality of a steel strip as described in claim 1, wherein the abnormal detecting step comprises: selecting a plurality of consecutive images from the steel strip images; The vertical coordinate values of the coordinates of the center point of the largest object are summed to obtain a total value of the vertical coordinates; the total value of the vertical coordinates is divided by the number of the consecutive images to obtain a vertical coordinate average value, and the vertical coordinate is used The average value is used as a vertical coordinate of a center line of the steel strip image; and an offset determining step is performed to determine whether the steel strip in the second steel strip image has a tail end offset according to the vertical coordinate of the center line Happening. The method for detecting a rolling end abnormality of the steel strip according to Item 7 of the claim, wherein the offset determining step comprises: selecting at least one image to be processed from the steel strip images, wherein the image to be processed a continuous image, and the second steel strip image is the last one of the at least one image to be processed; calculating a distance between the center coordinate of the largest object of each of the to-be-processed images and the center line to obtain a plurality of steels With trailing end offset value; Calculating a distance between the central coordinate of the largest object of each of the to-be-processed images and an upper edge of the circumscribed rectangle to obtain an amplitude of the upper edge of the plurality of steel strips; calculating the maximum of each of the to-be-processed images The distance between the central coordinate of the object and the lower edge of the circumscribed rectangle to obtain the amplitude of the lower edge of the plurality of steel strips; calculating the difference between the upper and lower limits of the maximum amplitude and the minimum of the upper edges of the strips And determining whether the upper edge oscillation amplitude difference is greater than an upper edge oscillation amplitude threshold to provide a first judgment result; calculating a lower edge oscillation amplitude of one of a maximum value and a minimum value of the lower edge oscillation amplitudes of the steel strips a difference, and determining whether the lower edge oscillation amplitude difference is greater than a lower edge oscillation amplitude threshold to provide a second determination result; and according to the first determination result, the second determination result, and the steel strip tail end offset The value is shifted to determine if the steel strip is offset. The method for detecting the abnormality of the end rolling of the steel strip according to Item 8 of the claim, wherein when the maximum value of the offset value of the tail ends of the steel strips is greater than a first threshold, the first judgment result is When the second judgment result is YES, it is judged that the tail end of the steel sheet is shifted upward. The method for detecting the abnormality of the end rolling of the steel strip according to Item 8 of the claim, wherein when the minimum value of the offset value of the tail ends of the steel strips is less than a second threshold, the first judgment result and When the second judgment result is YES, it is judged that the tail end of the steel sheet is shifted downward.