JPS61294307A - Detection for binding material of coil - Google Patents

Abstract

PURPOSE:To improve the detection precision of a hoop by illuminating a coil obliquely and detecting the brightness level as a picture signal to perform the edge emphasizing processing in the rolling direction of the coil. CONSTITUTION:A binding hoop 2 is set to a coil 1. A coil placing device 3 rotates the coil 1. Illuminating devices 4-1 and 4-2 irradiate the coil 1 obliquely to form the shadow due to the step between the outside peripheral surface of the coil 1 and the hoop 2. An ITV camera 5 picks up the image of the coil 1, and the picture signal is inputted to an AD converter 6. The signal from the converter 6 is stored in a storage part 8 of a storage arithmetic unit 7. An arithmetic part 9 uses a preliminarily determined edge emphasizing processing operator to detect the position of the hoop 2.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention provides a coil binding material 1, for example, a hoop around the outer circumference of the coil,
This relates to a method for automatically detecting a hoop and contributes to the automatic coil opening process of cutting the hoop and unwinding the coil.

[Conventional technology]

In general, the opening process of a coil is a process in which, for example, hoops that bind the outer circumference of a coil wound after hot rolling are removed in the primary process to release the coil. Unwrap it and put it into a pinch roll, for example.

In the conventional opening process, the operator looks at the hoop binding the coils and cuts the hoop with a cutting tool.6 When automating the coil opening process as described above, the hoop is cut. However, for this we first need to detect the position of the hoop.

[Problem that the invention seeks to solve]

By the way, if the coil is a coil after hot rolling, for example, two hoops may be provided depending on the plate thickness and the shape of the tip.

In addition, the coil shrinks when it is cooled after hooping, and the hoop may become loose and shift from the center or become slanted.Furthermore, the coil may be moved between the hot rolling line and the pickling line to make it ``inner vertical''. If the hoop is conveyed by the hoop, the misalignment of the hoop will further increase, and in extreme cases, the hoop may fall off.

The position of the hoop that binds the coils in this way is not necessarily in a fixed position, but rather often changes.
Therefore, to fully automate coil opening, it is necessary to automatically detect the hoop position.

An object of the present invention is to provide a method for automatically detecting the position of a coil binding material such as a hoop.

[Means for solving problems]

In the present invention, the coil is irradiated with a light beam from an oblique direction, and the intensity is created by the level difference on the surface of the coil.

This is imaged with a television camera, the image signal is stored on the screen in units of pixels divided into line segments in advance, and the stored image signal is calculated with a preset data pattern for step enhancement processing, and the coil The image signals existing in the rolling direction are weighted, and the calculated value in step 7 is compared with a predetermined width signal of the binding material to detect the binding material.

The present invention will be explained in detail below based on one embodiment with reference to the drawings.

In FIG. 1, reference numeral 1 denotes a coil, and a hoop 2 for binding is attached to the coil l. Reference numeral 3 denotes a coil mounting device, such as a cradle, which allows the coil 1 to be rotated as required.

4-1.4-2 is an illumination device which irradiates the coil 1 with a light beam from an oblique direction to form a shadow due to the difference in level between the outer peripheral surface of the coil 1 and the hoop 2. At this time, it is preferable to make this illumination extremely strong to cause halation in areas other than the shadows caused by the steps, and to capture the image with the camera described later.In this way, image data that almost looks like it is binarized can be obtained. Image processing becomes easier.

In addition, as shown in FIGS. 2a and 2b, the hoop 2
On the other hand, by switching illumination of the coil 1 from the left and right directions using the illumination device 4-1 and 4-2, clear image data with shadows can be recorded. That is, first, as shown in Fig. 2a, the illumination 4-1 is turned on, and the illumination from the left creates a shadow on the right side of the hoop 2. Lighting device from the left, like l! 4-1 is turned off, lighting device 4-2 from the right is turned on,
Create a shadow to the left of hoop 2.

A camera, for example, an ITV camera 5, takes an image of the coil 1 that is obliquely irradiated with light, and inputs the image signal to the AD converter 6. One screen of the image signal AD-converted by the AD converter 6 is divided into, for example, 256 x 256 sections vertically and horizontally (one unit is called a pixel), and the image signal at each pixel is adjusted to the brightness level of the image. Accordingly, a digital chill image signal of multiple stages, for example, 0 to 225 stages (8 bits) is provided. The digital image signal for one screen from the AD converter 6 is stored in the storage unit 8 of the storage/arithmetic unit W17 in units of pixels as shown in FIG. In Fig. 3, the data for the entire screen is called A, and the data for each subdivided pixel is represented by A (tt j), where l is the pixel position in the rolling direction of the coil, and j is the coil position. represents the pixel position in the width direction.

In order to inspect the position of the hoop 2 over the entire circumference of the coil 1, the coil is rotated by an angle corresponding to one screen on the coil mounter 3 and then stopped, and then data is recorded. Repeat this operation for the entire circumference of the coil.

Further, the camera 5 may not have a two-dimensional field of view, but may have a one-dimensional field of view in the coil width direction, such as a linear sensor camera. In this case, the coil 1 is rotated for one revolution at a constant angular velocity at an appropriate speed using the coil mounting device 3, and data is recorded during that time.

Next, a method of edge detection based on optimal fitting will be explained.

An operator (data matrix) 1 for edge enhancement processing, for example, an operator MO shown in FIG. A product-sum operation is performed between M7 and the stored digital image signal for one screen.

That is, a product-sum operation is performed between the numerical values determined by the operators MO to M7, which are 3×3 horizontally, and the pixel data cut out for 3×3 pixels from the digital image signal for one screen of 256×256 pixels. .

A concrete example of this method is as follows.

Figure 5 shows the data for each position of the 3x3 edge detection operator M, and the numerical value of each element (corresponding to pixel) is
t by...i. The digital image signal A to be processed is similar to that shown in FIG. Let B be the digital image signal after processing.

The image signal A is stored in the storage section 8, and the arithmetic section 9
The following equation is calculated in .

B (i, j)=A (i-1, j -L) Xa+
A (i-1,j) Xb+ A (i-1,j+1)
Xc+A(i, j-1) Xd+A(i, j) Xe
+A(i,j+L)Xf+A(i+1.j-1)Xg
+A(i+1.j)Xh+A(i+1.j+1) Xi
In this method, the above equations are calculated for each of the eight types of operators shown in FIG. 4, the eight sets of results are compared, and the largest value is stored as the "edge strength" at that location.

However, here, we want to emphasize only the signals that have edges in the direction parallel to the coil rolling direction, so the "edge direction" is set to M.
2. For "edge intensities" other than M6, this calculation is performed sequentially from the upper left of the matrix A of pixels of one screen, and is repeated while shifting each pixel to the right. When the right edge has been processed, it shifts down by one pixel and starts processing again from the left edge to the right.

When processing is completed to the bottom right, the entire screen has been processed.

This process is applied to two data items: data 1 (Figure 6) illuminated by illumination device 4-1 from the left in Figure 1, and data 2 (Figure 6) illuminated by illumination device 4-2 from the right. Do each separately.

Next, a method for synthesizing these two data 1 and 2 will be described.

The calculation unit 9 of the storage calculation unit 7 compares the corresponding pixels of the edge-enhanced data 1 and 2 stored in the storage unit 8, and selects the one with the larger edge strength value. This is displayed on the display device 10 as data 3 (FIG. 6).

Data 3 here represents the degree to which each pixel is an edge in the coil rolling direction. The stronger the degree, the larger the numerical value will be displayed in black on the display device.

Here, in order to simplify subsequent processing, binarization is performed in the arithmetic unit 9. Binarization here means that pixels whose degree of edge is equal to or higher than a preset level are recognized as edges. Let it be 1. Conversely, a pixel below a certain level is set to O.

Through the above processing, binary data 4 (FIG. 6) representing pixels having vertically continuous edges was obtained.

Next, feature extraction of hoop 2 is performed.

The hoop 2 has a constant width of, for example, 30 mm in this embodiment, and is a band-shaped thing corresponding to 19 pixels. This feature is
The number of hoops 2 does not change even if the hoops 2 are diagonal. Therefore, using this feature, extract the hoop 2 portion 6. In other words, out of data 4 (Figure 6), the width of hoop 2 is separated. Only one pair of data is extracted, and specifically, the calculation is performed as follows. Data 4 (Figure 6) is 2
The value of each element of 56 rows and 256 columns is represented by a matrix of 0 or 1. Let this matrix be C.

If C(i, j)XC(i, j+n-1)=1, then C(i, j)=1, C(i, j+n-1)=1
shall be.

If C(i, j) X C(i, j + n-1)
= When O.

Let C(i, j)=0 and C(i, j+n-1)=0.

However, ・n is the number of pixels equivalent to the hoop width.

In this example, 19, i represents the position in the coil rolling direction, and j represents the position in the coil width direction. By performing this correlation calculation for the width of the hoop, most of the noise on both sides of the coil and the one-shot noise can be removed, and data 5 (Fig. 6) consisting mostly of the edge portion of hoop 2 can be obtained. I can do it. At this stage, hoop 2 is extracted.

Next, the data 5 (FIG. 6) displayed on the display device 10 is projected in the coil width direction. This result indicates where the hoop edges, which are continuous in the coil rolling direction, are distributed in the width direction of the coil.

Specifically, the following calculations are performed.

The calculation result D(j) is shown in data 6 (FIG. 6).

By binarizing this at a certain level, the position in the width direction of the hoop 2 can be detected as a pixel number.

The rolling direction edge emphasis during the hoop 2 position detection algorithm is performed by the operator shown in FIG. 4, but the operator is not limited to this and may be used for other rolling direction edge emphasis. Another example is shown in FIG. 7a. and FIG. 7b).

〔Effect of the invention〕

As described above, the present invention illuminates the coil from an oblique direction, detects the brightness level as an image signal, and performs edge enhancement processing in the coil rolling direction to detect the hoop, so the detection accuracy is high and This can be added automatically.

This detected signal such as the hoop position is utilized as information to a separate coil opening device.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing an overview of the configuration of an apparatus that implements one embodiment of the present invention. FIGS. 2a and 2b are perspective views showing the illumination mode of the coil in one embodiment of the present invention. FIG. 3 is a diagram for explaining the storage state of image signals in the storage unit 8 shown in FIG. 1, and is a plan view showing a two-dimensional distribution of one image signal. FIG. 4 is a plan view showing the contents of an operator used in an embodiment of the present invention. FIG. 5 is a plan view showing data at each position within the operator. FIG. 6 is a plan view showing a two-dimensional distribution of image data obtained at each signal processing stage in an embodiment of the present invention. FIGS. 7a and 7b are plan views showing other examples of the operator. 1: Coil 2: Hoop 3: Cradle 4-1.4-2: Lighting device 5: Television camera 6: AD converter 7: Memory calculation device 8
: Storage unit 9: Calculation unit lO: Display device East 2b Figure East 3 Figure 4 To 5 Section 7a Figure East 7b

Claims (1)

[Claims] A light beam is irradiated onto the coil from an oblique direction to create a level of contrast on the surface of the coil, which is imaged with a television camera, and the image signal is stored in units of pixels, which are divided into line segments on the screen in advance. , calculates the stored image signal with a preset step emphasis data pattern, converts the step signal parallel to the coil rolling direction on the coil surface into an image signal that emphasizes it, and further calculates the width of the binding material, which is known in advance. 1. A method for detecting a coil binding material, characterized in that only a pair of step signals having an interval of 10 minutes is extracted and detected as the position of the binding material in the coil width direction.